diss deposit final ohnecv - research collection47855/et… · diss. eth no. 22589 engineered...

Research Collection

Doctoral Thesis

Engineered nanomaterials in the agricultural environmentCurrent state of applications and development of analyticalmethods

Author(s): Gogos, Alexander

Publication Date: 2015

Permanent Link: https://doi.org/10.3929/ethz-a-010473555

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.

ETH Library

https://doi.org/10.3929/ethz-a-010473555

http://rightsstatements.org/page/InC-NC/1.0/

https://www.research-collection.ethz.ch

https://www.research-collection.ethz.ch/terms-of-use

DISS. ETH NO. 22589

ENGINEERED NANOMATERIALS IN THE AGRICULTURAL ENVIRONMENT:

CURRENT STATE OF APPLICATIONS AND DEVELOPMENT OF ANALYTICAL METHODS

A thesis submitted to attain the degree of

DOCTOR OF SCIENCES of ETH ZURICH

(Dr. sc. ETH Zurich)

presented by

Alexander Gogos

Dipl.-Biol., RWTH Aachen University

born on 10.11.1983

citizen of Germany

accepted on the recommendation of

Prof. Renato Zenobi (examiner) Prof. Konrad Hungerbühler (co-examiner)

Dr. Thomas D. Bucheli (co-examiner)

2015

______________________________________________________________________________________

„Μέτρον ἄριστον“

-Κλεόβουλος-

To my parents

Antonios and Christa

iii

Acknowledgements

First of all I would like to thank Prof. Renato Zenobi for giving me the chance to

do my PhD under his supervision. I am very grateful for the scientific freedom,

advice and support he has given me during the entire time. Second - but no less -

, I would like to express my gratitude to Dr. Thomas Bucheli who made this pro-

ject possible and who was always there when he was needed! Thomas, you have

been a great “boss”! I also like to thank Prof. Dr. Konrad Hungerbühler for agree-

ing on being my co-examiner.

I like to thank Janine Moll for being such a great company during this project and

for inaugurating me into the many secrets of the Swiss culture and language, es-

pecially the “Oute”-dialect (let’s do an OECD Ouge-Tescht!).

Thanks go also to the rest of the “NANOMICROPS”-Team, Dr. Katja Knauer, Dr.

Franco Widmer, and Prof. Dr. Marcel van der Heijden.

Furthermore, I am grateful for so many amazing colleagues within the Agroscope

Analytical Chemistry group that made these nearly four years really pleasant:

Especially Dr. Hans-Jörg Bachmann, who made a lot possible “behind the

scenes”, Dr. Felix Wettstein for his good company in the office and the many dis-

cussions about all kinds of topics, Franziska Blum and Dr. Hans Stünzi for their

friendship and for introducing me into the secret world of caves, Dr. Isabel -

“three zältli at a time” - Hilber for her support with R and fun in the lab, Diane

Bürge for her support with the ICPs and Nora Bartolomé for enduring my “kölsche

Dialekt”. Not to forget the many other people in this group but also at Agroscope

in general that contributed a lot – visible or invisible – to the nice atmosphere but

also to make a lot of things easier.

Acknowledgments

iv

Special thanks go also to Dr. Xanat Flores-Cervantes, who introduced me to the

world of the Nanotubes, and her husband Gregory Tkac - you both have been a

great “multi-kulti” company here.

Furthermore I enjoyed working with many collaborating partners: I am very thank-

ful to Dr. Ralf Kägi for his unbreakable enthusiasm (contaminations begone!) and

the many vital contributions to our project, as well as Brian Sinnet for harboring

and supporting me in the EAWAG particle lab.

I also like to express my gratitude to Prof. Dr. Vera Slaveykova and Dr. Monika

Mortimer for the pleasant and fruitful collaboration on hyperspectral imaging. Fur-

thermore, I was lucky to get into contact with Prof. Dr. Micah Green and Fahmida

Irin, who contributed an important part to the nanotube exposure assessment,

thank you!

Thanks go also to the team at ZMB Zurich, especially Dr. Andres Kaech who

taught me a lot about electron microscopy and inspired me with excitement for

this technique, as well as Ursula Lüthi and Dr. Bruno Guhl for their lab support.

Furthermore, I like to thank Dr. Alexandra Kroll for giving us the possibility to do

some nano-tracking analysis measurements at EAWAG.

I thank the team at Cytoviva, especially Dr. Jim Beach, as well as the team at

Postnova and especially Dr. Thorsten Klein for their technical support regarding

hyperspectral imaging and field-flow fractionation, respectively.

I thank Dr. Nataliya Fedotova for bringing an interesting project and a more than

equally interesting device to our lab. I really enjoyed this time.

Furthermore, I want to thank some people “behind the scenes”: Dr. Franziska

Schwabe, Dr. Johannes Pfeifer, Dr. Michael Mielewczik, Dr. Michael Evangelou,

Björn Studer and Dr. Jens Botterweck for their friendship and all the good times

we had!

I want to thank all my Greek and German family for their constant belief in me,

and last but not least, the ones that probably contributed the most to everything:

my parents Antonios and Christa and my love Chantal. Without you, I would not

be here.

v

Abstract

Nanotechnology has undergone enormous developments in the last decade and

has raised high expectations in many industrial fields. There is already scientific

certainty that the production and the use of nanomaterials (NMs) leads to their

release into the environment via different pathways. A dramatic increase of NM

fluxes into the environment could be expected, if NMs were released intentional-

ly. This might be the case for example in agriculture, where they are foreseen as

ingredients that enhance the effectiveness and sustainability of plant protection

products and fertilizers. Despite these advances and benefits that may be deliv-

ered by nanotechnology, there is a growing concern about the environmental

safety of NMs for the last decade. Several adverse effects attributed to the pres-

ence of NMs have been reported for microorganisms, plants, animals and hu-

mans. Therefore, to prevent potential harm arising from future (agricultural) NM

applications, there is a need to - proactively – generate tools that allow assessing

their risk in the environment. A first prerequisite for such a risk assessment is

knowledge on the potential NM exposure. With regard to agriculture, this means

knowledge on the current state of research, patents and actual nano-plant protec-

tion or fertilizer products on the market. However, a systematic compilation of the

developments in this particular field as well as of actual applications potentially

already on the market was lacking at the beginning of this thesis. Therefore the

literature was compiled and critically reviewed. It could be shown that scientific

publications and patents on NMs for use in plant protection or fertilizer products

have exponentially increased since the millennium shift. However, the dynamic

development in research and its considerable public perception were in contrast

to the currently still very small number of NM-containing products on the market.

In addition, the absolute number of scientific papers and patents were rather low

compared to other more prominent fields, such as energy and electronics. A sec-

ond prerequisite for environmental risk assessment of NMs is the ability to de-

Abstract

vi

scribe the actual exposure using analytical methods. However, due to the enor-

mous complexity of environmental matrices, the detection and quantification of

NMs in such systems is very challenging and needs a complex analytical chain.

This thesis therefore aimed at contributing to novel methods within the analytical

chain in increasing stages of complexity. In a first step, the potential of hyper-

spectral imaging with enhanced darkfield microscopy (HSI-M) was investigated

for its capabilities to examine the cellular uptake of different metal-based nano-

particles (NPs) into the protozoon Tetrahymena thermophila within a rather sim-

ple aqueous test system. HSI-M was chosen as it possesses unique advantages

due to its simple and quick sample preparation and non-invasiveness. It could be

shown that the chemical specificity of the technique is probably limited to metallic

NMs that exhibit localized surface plasmon resonance. However, the technique

was also applicable to particles with an unspecific spectral behaviour, provided

that the spectral background of the matrix was different. HSI-M was also able to

differentiate between different NP types (e.g., between silver NPs and cadmium-

selenide quantum dots), depending on their spectral profiles. It could be shown

that HSI-M could in principle be used to quantify and localize NMs taken up into

the cells. However, the results remained on a semi-quantitative level, as the

measured parameter (matched pixels per cell area) could not be related to an

absolute particle mass concentration in the course of the study. In addition, the

currently long analysis times hinder the acquisition of statistically strong datasets.

Apart from aqueous systems such as surface waters, (agricultural) soils also rep-

resent a potential compartment that can receive the increasing amounts of differ-

ent nanomaterials that enter the environment. Carbon nanotubes (CNTs) repre-

sent an important class of NMs that are among the most produced and used

worldwide. Analytical detection and quantification of these NMs in soils is very

challenging due to the enormous amount of background carbon. While labile or-

ganic compounds as well as carbonates can be easily removed, refractory car-

bon such as natural and anthropogenic soot still interferes in elemental analysis.

Therefore, a novel method was developed that made use of shape differences

between the analyte and background particles. The difference in shape could be

described by a shape factor ρ that was derived from light scattering measure-

ments from pre-fractionated samples. This allowed to specifically detect multi-

walled CNTs in soil samples with detection limits between 1.6 and 4 mg g-1.

Abstract

vii

These detection limits were, however, much higher than any currently predicted

environmental concentration of these NMs. The results of the light scattering

measurements with pure multi-walled CNT suspensions could additionally be

confirmed using automated electron microscopy (EM) and image analysis. In a

small initial approach, it could also be shown that in principle, automated EM im-

age analysis has the potential to generate reasonable number based concentra-

tions of pure multi-walled CNT suspensions. This approach may be transferred to

real soil samples in the future, with the advantage of increased sensitivity and

simultaneous acquisition of both particle number and distributions of their dimen-

sions. In addition, if the sample preparation can be further optimized, analysis

times could be fast enough for routine use. Before this can be achieved, howev-

er, the level of organic contaminations has to be further reduced and the image

analysis procedure optimized. Most risk assessment studies until now have dealt

with artificial systems (e.g., aqueous cultures). Soil, as one important sink for

NMs has however received less attention in this respect. Therefore, in a com-

bined exposure and effect study, an elaborate pot experiment was conducted

with agriculturally important plants (clover and wheat). The plants were exposed

to titanium dioxide nanoparticles (TiO2 NPs) and multi-walled CNTs which were

homogenously mixed into the soil. While in the companion study the effects of the

NMs on the plants and symbiotic organisms were investigated, here, a battery of

analytical techniques was applied to rigorously confirm the actual exposure con-

centrations and to describe vertical NM translocation in the soil and plant uptake.

It could be shown that both types of NMs exhibited a rather limited mobility in the

soil. While no significant plant uptake could be observed for TiO2 NPs, some clo-

ver plants took up multi-walled CNTs from the soil, independent of the exposure

concentration.

ix

Zusammenfassung

Die Nanotechnologie hat sich seit der Jahrtausendwende rasant entwickelt und

weckt dadurch grosse Erwartungen, vor allem in der Industrie. Es gilt mittlerweile

als wissenschaftlich erwiesen, dass die Produktion und die Nutzung von Produk-

ten, welche Nanomaterialien (NM) enthalten, dazu führt, dass diese über ver-

schiedene Wege in die Umwelt gelangen. Einen weitaus höheren als den bisher

vorausgesagten Grad der Freisetzung solcher NM kann man allerdings erwarten,

falls solche Materialien absichtlich in die Umwelt ausgebracht werden. Dies könn-

te beispielsweise in der Landwirtschaft der Fall sein. Dort erhofft man sich durch

den Zusatz von NM eine effektivere und nachhaltigere Nutzung von Pflanzen-

schutzmitteln und Düngern. Allerdings gibt es neben der Wahrnehmung der Vor-

teile und (Produkt-)Verbesserungen durch NM in den letzten Jahren auch zu-

nehmend Bedenken, was ihre Umweltverträglichkeit betrifft. Diese Bedenken

werden durch Studien gestützt, die zeigen, dass durch NM negative Effekte bei

Mikroorganismen, Pflanzen, Tieren und Menschen ausgelöst werden können.

Um mögliche Umweltauswirkungen abschätzen zu können ist es notwendig,

rechtzeitig das nötige „Handwerkszeug“ bereitzustellen, welches eine Risikoana-

lyse solcher NM in der Umwelt erlaubt. Eine erste Voraussetzung für eine solche

Risikoanalyse ist zunächst einmal das Wissen um die potentielle Exposition. Das

heißt konkret: man muss den aktuellen Stand bezüglich Forschung und Entwick-

lung einerseits, und den aktuellen Stand marktreifer oder auf dem Markt befindli-

cher Produkte andererseits kennen. Allerdings fehlte zu Beginn dieser Arbeit eine

systematische Zusammenstellung dieser dringend benötigten Fakten. Daher

wurde als Einstieg in das Thema die bestehende Literatur gesichtet und kritisch

betrachtet. Es konnte gezeigt werden, dass die Zahl an wissenschaftlichen Veröf-

fentlichungen und Patenten, die sich mit der Nutzung von NM in Pflanzen-

schutzmitteln und Düngern befassen, seit der Jahrtausendwende exponentiell

zugenommen hat. Allerdings standen diese dynamischen Entwicklungen inner-

Zusammenfassung

x

halb der Forschung sowie die bemerkenswerte öffentliche Wahrnehmung stark

im Kontrast zu der eher kleinen Anzahl an NM-enthaltenden Produkten, die sich

zu diesem Zeitpunkt nachweislich auf dem Markt befanden. Weiterhin war auch

die absolute Anzahl an Veröffentlichungen und Patenten viel niedriger als in an-

deren Nanotechnologie-Sparten, wie beispielsweise im Energie- und Elektronik-

bereich. Eine weitere Voraussetzung für eine Umweltrisikobeurteilung von NM ist

die Fähigkeit, die tatsächliche Exposition mittels analytischer Methoden zu quan-

tifizieren. Allerdings ist diese Aufgabe aufgrund der enormen Komplexität von

Umweltproben, wie beispielsweise Boden, eine sehr große Herausforderung, die

eine Kette von verschiedenen analytischen Herangehensweisen benötigt. Daher

ist es ein Ziel dieser Arbeit zu neuen analytischen Methoden innerhalb dieser

Kette in aufsteigender Komplexität beizutragen. In einem ersten Schritt wurde in

einem relativ einfachen wässrigen System die hyperspektrale Bildgebung gekop-

pelt mit verstärkter Dunkelfeldmikroskopie (HSI-M) auf ihre Fähigkeit hin unter-

sucht, die Aufnahme verschiedener metallischer NM in Zellen von Tetrahymena

thermophila zu untersuchen und zu quantifizieren. Der enorme Vorteil den diese

Technik von vorherein bot, war die einfache und schnelle Probenvorbereitung

sowie die Möglichkeit, das Verhalten der NM nicht-invasiv zu beobachten. Es

zeigte sich, dass die chemische Spezifität der Methode vermutlich auf solche me-

tallische Partikel beschränkt ist, welche eine Oberflächenplasmonresonanz zei-

gen. Allerdings konnte die Methode auch auf Partikel mit weniger spezifischem

spektralem Verhalten angewendet werden, vorausgesetzt, dass sie sich in ihrem

Spektrum deutlich von der Hintergrundmatrix unterschieden. Ebenfalls abhängig

von ihrem spektralen Verhalten, konnten auch verschiedene Typen von NM von-

einander unterschieden werden (z.B. Silbernanopartikel von Cadmium-Selenit

Quantum dots). Grundsätzlich konnte HSI-M gut genutzt werden um Partikel in-

nerhalb der Zellen zu lokalisieren und ihre Aufnahme zu quantifizieren. Allerdings

blieben die Resultate auf einem semi-quantitativen Niveau, da es innerhalb die-

ser Studie nicht gelang, die notwendige Beziehung zwischen dem gemessenen

Parameter (NM-Pixel pro Zellfläche) hin zu einer absoluten massenbasierten

Konzentration der Partikel herzustellen. Der momentan noch relativ hohe Zeitbe-

darf pro Messung erschwert zudem die Aufnahme einer statistisch relevanten

Anzahl an Datenpunkten.

Zusammenfassung

xi

Neben den verschiedenen wässrigen Umweltsystemen, wie z.B. Oberflächenge-

wässern, stellen Böden ebenfalls ein wichtiges Kompartiment dar, in das NM ge-

langen können. Der Nachweis und die Quantifizierung von Kohlenstoffnanoröh-

ren (CNTs) - einer wichtigen Klasse von NM, die zu den am meisten produzierten

und genutzten Partikeln weltweit zählt - ist eine sehr große Herausforderung in

Böden, da diese Matrix einen vergleichsweise hohen Anteil an Kohlenstoffhinter-

grund aufweist. Während man labile organische Verbindungen sowie Karbonate

relativ einfach entfernen kann, verbleiben hitzebeständige Kohlenstoffverbindun-

gen wie zum Beispiel natürlicher bzw. anthropogener Ruß im Boden und stören

in der folgenden Elementaranalyse. Daher wurde eine neuartige Methode zur

Detektion entwickelt, die sich den Formunterschied zwischen den CNTs und den

Hintergrundpartikeln zu Nutze macht. Dieser Formunterschied wird durch einen

Formfaktor ρ quantifiziert, der von Lichtstreumessungen an einer vorgängig frak-

tionierten Probe abgeleitet wird. Dies erlaubte es, CNTs spezifisch in Bodenpro-

ben nachzuweisen. Die Detektionsgrenzen lagen hierbei zwischen 1.6 und 4 mg

g-1, was allerdings um ein vielfaches höher ist als jede bisher vorhergesagte

Umweltkonzentration. Die Ergebnisse der Lichtstreumessungen an reinen Sus-

pensionen von mehrwandigen CNTs konnten mithilfe von automatisierter Elekt-

ronenmikroskopie (EM) und nachgeschalteter Bildanalyse orthogonal bestätigt

werden. In einem ersten Ansatz konnte zudem auch gezeigt werden, dass die

automatisierte EM zusammen mit der Bildanalyse in der Lage ist, sinnvolle an-

zahlbasierte Konzentrationen von reinen CNT-Suspensionen zu liefern. Dieser

Ansatz hat das Potenzial in Zukunft auch auf Bodenproben angewendet werden

zu können. Zudem könnte er den Vorteil einer erhöhten Empfindlichkeit bieten

und es erlauben, gleichzeitig Informationen über die Zahl der Partikel sowie über

ihre Dimensionen zu erhalten. Sollte die Probenpräparation noch weiter optimiert

werden können, so könnten die Analysezeiten sogar für einen Routinebetrieb

geeignet sein. Bevor dies allerdings erreicht werden kann, muss die Menge an

auftretenden organischen Kontaminationen dringend reduziert, sowie der Bild-

verarbeitungsprozess weiter optimiert werden. Die meisten Studien zur Risikobe-

urteilung von NM haben sich bisher hauptsächlich mit künstlichen Systemen, wie

z.B. Hydrokulturen, befasst. Dem Boden als wichtige Senke für NM wurde dabei

bisher noch zu wenig Bedeutung beigemessen. Daher wurde in einer kombinier-

ten Effekt- und Expositionsstudie ein aufwendiges Topfexperiment mit für die

Zusammenfassung

xii

Landwirtschaft wichtigen Pflanzenarten (Klee und Weizen) durchgeführt. Die

Pflanzen wurden hierbei in Böden kultiviert, in die jeweils vorgängig Titan-dioxid

Nanopartikel (TiO2-NP) oder mehrwandige CNTs homogen eingearbeitet worden

waren. Während in der Partnerstudie die Effekte der NM auf die Pflanzen sowie

symbiotische Mikroorganismen untersucht wurden, wurde hier eine Reihe von

analytischen Methoden dazu eingesetzt, die tatsächlichen Expositionskonzentra-

tionen zu bestätigen und zudem die Verlagerung der NM im Boden und deren

Aufnahme in Pflanzen zu beschreiben. Es konnte gezeigt werden, dass beide

untersuchte Arten von NM im Boden eher beschränkt mobil waren und es keine

signifikante Aufnahme von TiO2-NP in die Pflanzen gab. Allerdings wurden, un-

abhängig von der in den Boden eingesetzten Konzentration, mehrwandige CNTs

aus dem Boden in Kleepflanzen aufgenommen.

xiii

Contents

Acknowledgements ...................................................................................................................................................... iii

Abstract ........................................................................................................................................................................... v

Zusammenfassung ....................................................................................................................................................... ix

Contents ....................................................................................................................................................................... xiii

Abbreviations ............................................................................................................................................................. xvii

General introduction ............................................................................................................................ 19 Chapter 1

1.1 Nanotechnology and nanomaterials ........................................................................................................... 19

1.2 Types, properties and use of nanomaterials ............................................................................................. 21

1.3 Release of nanomaterials into the environment ........................................................................................ 23

1.4 Environmental risk assessment of nanomaterials .................................................................................... 25

1.5 Main nanomaterials investigated in NANOMICROPS.............................................................................. 26

1.5.1 Titanium dioxide ................................................................................................................................ 26

1.5.2 Carbon nanotubes ............................................................................................................................ 27

1.6 Challenges and analytical approaches for NM characterization and quantification in complex matrices .......................................................................................................................................................... 29

1.6.1 General analytical challenges ......................................................................................................... 29

1.6.2 Sample preparation and storage .................................................................................................... 31

1.6.3 Analytical strategies: The need for sophisticated analytical chains ........................................... 32

1.6.4 Specific challenges and approaches for the analysis of TiO2 NPs in soils ............................... 33

1.6.5 Specific challenges and approaches for the analysis of CNTs in soils ..................................... 34

1.7 Main analytical techniques employed in this thesis .................................................................................. 36

1.7.1 Hyperspectral imaging microscopy ................................................................................................ 36

1.7.2 Asymmetric flow field-flow fractionation (aF4) .............................................................................. 38

1.7.3 Multi-angle light scattering (MALS) detection ............................................................................... 43

1.8 Objectives and contents of the thesis within the NRP64 and the project NANOMICROPS ............... 48

Nanomaterials in plant protection and fertilization : current state, foreseen applications Chapter 2and research priorities .................................................................................................................................... 51

Abstract ....................................................................................................................................................................... 52

2.1 Introduction ..................................................................................................................................................... 53

2.2 Methodology ................................................................................................................................................... 55

2.3 Evolution of research and development activities for NMs in agriculture .............................................. 57

Contents

xiv

2.4 Classes of NMs and their intended purpose in agriculture ...................................................................... 60

2.4.1 Non-solid NMs ................................................................................................................................... 62

2.4.2 Solid NMs ........................................................................................................................................... 66

2.5 Research priorities for a safe use of NMs in PPP and fertilization ......................................................... 74

2.6 Nano-regulation and on-going activities ..................................................................................................... 76

2.7 Acknowledgements ....................................................................................................................................... 77

2.8 Supporting information ................................................................................................................................. 78

Potential of Hyperspectral Imaging Microscopy for se mi-quantitative analysis of Chapter 3nanoparticle uptake by protozoa .................................................................................................................. 87

Abstract ....................................................................................................................................................................... 88

3.1 Introduction .................................................................................................................................................... 89

3.2 Experimental .................................................................................................................................................. 91

3.2.1 Nanoparticles .................................................................................................................................... 91

3.2.2 Protozoan culture .............................................................................................................................. 92

3.2.3 Exposure of protozoa to NPs and preparation of the cells for microscopy ............................... 92

3.2.4 Preparation of reference samples .................................................................................................. 93

3.2.5 Hyperspectral imaging ..................................................................................................................... 94

3.2.6 Image processing and analysis ...................................................................................................... 94

3.2.7 Statistical analysis ............................................................................................................................ 95

3.3 Results and discussion ................................................................................................................................. 96

3.3.1 Influence of protozoan extracellular substances (ES) on visible and near infrared (VNIR) spectral profiles of nanoparticles and the characteristics thereof ................................. 96

3.3.2 Spectral angle mapping of nanoparticles in protozoa .................................................................. 98

3.3.3 Nanoparticle differentiation capabilities of HSI-M ...................................................................... 100

3.3.4 Semiquantitative characterization of nanoparticle uptake and clearance by T. thermophila as measured by HSI-M ............................................................................................. 102

3.4 Application potential of HSI-M and outlook .............................................................................................. 106

3.5 Acknowledgements ..................................................................................................................................... 106

3.6 Supporting information ............................................................................................................................... 108

3.6.1 Experimental .................................................................................................................................... 108

Capabilities of asymmetric flow field-flow fraction ation coupled to multi-angle light Chapter 4scattering to detect carbon nanotubes in soot and s oil .......................................................................... 121

Abstract ..................................................................................................................................................................... 122

4.1 Introduction .................................................................................................................................................. 123

4.1.1 Theory .............................................................................................................................................. 124

4.2 Materials and methods ............................................................................................................................... 125

4.2.1 Chemicals and analytes ................................................................................................................. 125

4.2.2 Soils and sediments ....................................................................................................................... 126

4.2.3 Analyte suspensions and sample extracts .................................................................................. 126

Contents

xv

4.3 Results and discussion ............................................................................................................................... 131

4.3.1 General features of aF4-MALS fractograms and consequences for their interpretation ...... 131

4.3.2 aF4-MALS analysis of MWCNTs and MW1-soot mixtures ....................................................... 132

4.3.3 Orthogonal confirmation of aF4-MALS results using automated EM in combination with image analysis ................................................................................................................................. 134

4.3.4 aF4-MALS analysis of native soils ............................................................................................... 136

4.3.5 aF4-MALS analysis of MW1 and soot spiked to soil extracts ................................................... 137

4.3.6 aF4-MALS analysis of MW1 and soot spiked to soils ................................................................ 139

4.3.7 Quality control and method validation .......................................................................................... 140

4.3.8 Further optimization of aF4-MALS ............................................................................................... 141

4.3.9 aF4-MALS within the larger analytical workflow ......................................................................... 142

4.4 Conclusions .................................................................................................................................................. 142



4.6.1 Example for the integration of aF4-MALS into a different analytical workflow: CTO-375 ..... 151

An initial approach to obtain number based concentr ations of CNTs using automated Chapter 5electron microscopy image analysis .......................................................................................................... 155

5.1 Motivation ..................................................................................................................................................... 155


5.2.1 Nanoparticles, dispersions and extracts ...................................................................................... 156

5.2.2 Sample preparation for automated EM ........................................................................................ 156

5.2.3 Automated EM analysis ................................................................................................................. 157

5.2.4 Image processing ............................................................................................................................ 157

5.2.5 Calculations ..................................................................................................................................... 158


5.3.1 Potential for the optimization of sample preparation and image processing .......................... 160

5.3.2 Quantitative evaluation ................................................................................................................... 161

5.3.3 First images from extracts of spiked soils ................................................................................... 163

5.3.4 Conclusion ....................................................................................................................................... 165

Titanium dioxide nanoparticles and carbon nanotubes in a soil mesocosm: Chapter 6vertical translocation in soil and plant uptake .......................................................................................... 167

Abstract ..................................................................................................................................................................... 168

6.1 Introduction ................................................................................................................................................... 169


6.2.1 Chemicals and nanoparticles ........................................................................................................ 170

6.2.2 Soil .................................................................................................................................................... 171

6.2.3 Spiking of the soil with NPs ........................................................................................................... 171

6.2.4 General experimental design ........................................................................................................ 172

Contents

xvi

6.2.5 Sampling of soil cores .................................................................................................................... 173

6.2.6 Titanium analysis in soils with XRF .............................................................................................. 173

6.2.7 Titanium analysis in leachates with ICP-OES............................................................................. 173

6.2.8 MWCNT quantification in soil with CTO-375 .............................................................................. 173

6.2.9 MWCNT analysis of soil with aF4-MALS ..................................................................................... 174

6.2.10 Titanium analysis of plants with ICP-OES ................................................................................... 174

6.2.11 MWCNT analysis of plants with MIH ............................................................................................ 175

6.2.12 Transmission electron microscopy of root cross sections ......................................................... 176

6.2.13 Statistics ........................................................................................................................................... 176


6.3.1 Vertical soil distribution and leaching of Ti .................................................................................. 176

6.3.2 Vertical soil distribution of BC/MWCNTs ..................................................................................... 179

6.3.3 Plant uptake of Ti ............................................................................................................................ 182

6.3.4 Plant uptake of MWCNTs .............................................................................................................. 185

6.4 Conclusions .................................................................................................................................................. 188



6.6.1 Detailed sample preparation steps of root cross sections for analysis using transmission electron microscopy ................................................................................................ 192

Conclusions and outlook .................................................................................................................. 193 Chapter 7

7.1 Conclusions .................................................................................................................................................. 193

7.2 Outlook.......................................................................................................................................................... 197

References .................................................................................................................................................................. 201

xvii

Abbreviations

Nanoparticles

Ag NP Silver nanoparticles

Au NP Gold nanoparticles

(MW, SW)CNT (Multi-walled, Single-walled) Carbon nanotubes

CuO NP Copper oxide nanoparticles

QDs Quantum dots

TiO2 NP Titanium dioxide nanoparticles

Analytical techniques

aF4 Asymmetric flow field-flow fractionation

BF Bright field (microscopy)

CTO-375 Chemo-thermal oxidation at 375 °C

EDX Energy dispersive X-ray spectroscopy

EM Electron microscopy

FFF Field-flow fractionation

HSI-M Hyperspectral imaging microscopy

ICP-OES Inductively coupled plasma - optical emission spec-troscopy

ICP-MS Inductively coupled plasma - mass spectrometry

MALS Multi-angle light scattering

MIH Microwave induced heating

SAM Spectral angle mapping

Abbreviations

xviii

SEM Scanning electron microscopy

STEM Scanning transmission electron microscopy

TEM Transmission electron microscopy

UV-Vis Ultraviolet-visible spectrophotometry

VNIR Visible to near infrared

XRF X-ray fluorescence spectroscopy

Other terms

BC Black carbon

IDL Instrument detection limit

IEP Isoelectric point

NM Nanomaterial

NP Nanoparticle

MDL Method detection limit

PES Polyether sulfone

PPP Plant protection product

PVDF Polyvinylidene fluoride

RC Regenerated cellulose

ROS Reactive oxygen species

RSD Relative standard deviation

SL Spectral library

19

General introduction Chapter 1

1.1 Nanotechnology and nanomaterials

In an etymological approach to unravel the term “nanotechnology” one could

begin with the prefix “nano”, which originates from the Greek word "νáνος“,

meaning “dwarf”. In combination with the likewise Greek suffix “technology”, orig-

inating from “τέχνη“, meaning an “art” or the “skill of hand” one can deduce that

“nanotechnology” might be the art of manipulation at a small scale. This is al-

ready quite similar to current definitions [1-3] that commonly regard nanotechnol-

ogy as the ability of manipulating matter at the “nano-scale” (i.e., 1-100 nm), at

which new material properties emerge that are not present in the materials bulk

counterpart and that are then explored in an interdisciplinary venture to improve

applications in various other technological fields. While the novelty and the use –

both scientifically and politically - of novelty claims regarding nanotechnology can

be argued to some extent [4], undoubtedly several milestones heralded the ad-

vent of a modern nanotechnology; some visionary, such as Feynman’s famous

talk at CalTech [5] or some technical, such as the invention of scanning tunneling

microscopy [6]. Since then, nanotechnological advances have undergone an

enormous development, which increased its pace dramatically after the year

2000/2001, when the National Nanotechnology Initiative was launched in the

United States [7]. A cumulative investment of 21 billion USD was carried out in

this governmental research initiative, supporting fundamental research as well as

such in the fields of health, energy and defense. The following rapid development

is reflected in the global market for nanotechnology containing products, which

was worth 254 billion US-Dollars (USD) in 2009 and is estimated to reach 3 tril-

lion USD by 2020 [7].

Despite the advances and benefits that nanotechnology may offer, there has also

been a growing concern in the last decade about the safety of the “building-

Chapter 1

20

blocks” of nanotechnology: nanomaterials (NMs). This public and also scientific

concern is based on the new nanoscale material properties (e.g., enhanced reac-

tivity) and is reflected in the apparent difficulty that regulators and scientists face

in defining such materials. When we talk about nanoscale materials usually -by

physical definition- we mean particulate entities with at least one external dimen-

sion between 1-100 nm. However, other operational definitions may be applied,

such as the latest definition by the European Commission [8]. Therein, a NM is

defined as “a natural, incidental or manufactured material containing particles, in

an unbound state or as an aggregate or as an agglomerate and where, for 50 %

or more of the particles in the number size distribution, one or more external di-

mensions is in the size range 1 nm - 100 nm. In specific cases and where war-

ranted by concerns for the environment, health, safety or competitiveness the

number size distribution threshold of 50 % may be replaced by a threshold be-

tween 1 and 50 %. By derogation from the above, fullerenes, graphene flakes

and single wall carbon nanotubes with one or more external dimensions below 1

nm should be considered as nanomaterials.” [8] Other definitions suggest to rely

on new (material specific) properties that emerge after the particles have passed

a certain size threshold [9] or further size-related properties such as the volume

specific surface area [10]. Such a yearning for a “one-size-fits-all”-definition is

mainly driven by the need for legal certainty regarding the regulation of NM con-

taining products which must be, in its essence, based on clear definitions [11].

This is criticized by some, who argue that other parameters than size and surface

area, such as for example shape, may govern detrimental effects, so that a “one-

size-fits-all” definition is doomed to fail [12]. As this discussion is still ongoing, this

thesis will allude to the –currently- most accepted definition by the European

Commission [8].

General introduction

21

1.2 Types, properties and use of nanomaterials

Recalling the first sentence in the EU definition of a NM, we notice that a NM can

also be “natural” [8]. This is very important, since it is a common misperception

that NMs are a new phenomenon based on advanced technology. They may oc-

cur in soils as “nanominerals” [13], such as e.g., goethite (α-FeO(OH)), and even

- at a first glance - exotic NM such as fullerenes have been found in carbon rich

rock [14], but even more surprisingly also on meteorites [15]. Many more of these

natural occurrences have been compiled by Nowack and Bucheli [16]. Nano-

materials also have been produced long before the advent of modern nanotech-

nology. Medieval artisans already crafted decorative films (e.g., in pottery) by re-

ducing Copper (Cu) and Silver (Ag) compounds in a controlled environment, re-

sulting in thin films containing spherical Cu and Ag nano-crystals [17]. Hundred

and twenty five years ago, the synthesis of a citrate-stabilized Silver-colloid was

reported and similar “colloidal” or “nano” Silver has been commercially used un-

der the name of “collargol” for medical applications since 1897 [18]. As plenty of

NMs occur naturally, and as many nanostructures have been produced long be-

fore modern nanotechnology, some scientists have raised the question, whether

contact to NMs represents a new type of hazardous exposure. In other words:

have organisms been exposed to such materials since the beginning of time

[19]? If this would be true, potential hazards to ecosystems would be less likely

and natural NM counterparts could serve as a blueprint for new but safe NMs

[19]. This discussion is still lively ongoing and probably cannot be answered

simply across-the-board due to the enormous diversity of NMs. Also, what has

changed nowadays is the amount in which NMs are produced and used, as well

as the speed in which new types of NMs are developed. For this reason, in the

following, the term “nanomaterial” or “nanoparticle” will always be used in refer-

ence to engineered or manufactured NMs.

The common goal of engineering a new NM is the enhancement of existing tech-

nologies or the creation of a completely new technology. This is possible because

of the new material properties that emerge at the nano-scale. Properties that

change once a particle of a material reaches the nano-size include mechanical,

optical, magnetic, electronic, thermal and catalytical ones [20]. The reason for the

difference in properties lies in interatomic interactions and an increased surface-

Chapter 1

22

to-volume ratio [20]. In products that make use of these properties, NMs can be

used in single-phase form (e.g., powders of a single NM type), multi-phase solids

(e.g., NMs embedded in composite materials, coated NMs) or multi-phase sys-

tems (e.g., colloids, heterogeneous dispersions, ferro-fluids) [21]. They can be

agglomerated, i.e. reversibly associated with either the same type of NM

(homoagglomeration) or with other particles (heteroagglomeration) [21]. Irreversi-

ble associations of NMs are termed aggregates.

Already now, nano-enhanced consumer products are available on the market in

significant quantities. For example, the public database “project on emerging

nanotechnologies” [22] has listed 1806 products, of which 437 are claimed to

contain Ag, 105 titanium (mostly as titanium dioxide, TiO2), 78 carbon (e.g., as

carbon nanotubes (CNTs)), 39 silicon and 38 zinc oxide (ZnO). However, such

numbers from a public database do not necessarily reflect actual production

quantities. Piccinno et al. [23] provided estimated values for global production

quantities of several NMs (median values). Silicon dioxide was highest with 5500

tons per year, followed by TiO2 (3000), ZnO (550) and CNTs (300). As the devel-

opment continues, also the production increases. For example the production

capacity for CNTs according to a more recent study [24] was 2300 Tons per year

(confirmed) and 4600 Tons per year (estimated) already in 2011.


23

1.3 Release of nanomaterials into the environment

Today, there is scientific certainty that produced and used NMs are also released

to the environment via different pathways [25] (see also Figure 1.1).

Figure 1.1: Routes of NMs to different environmenta l compartments leading to the expo-

sure of microorganisms, plants, animals and humans. Modified from [16].

The first – and certainly most important – type of release is unintentional and

arises from the mere use of NM containing products. In general, release can oc-

cur during production, transport, storage, use and disposal of NMs and NM-

containing products [26]. For example, Kaegi et al. [27] could show that nano-

TiO2 was released from exterior façade paints into surface waters. Several stud-

ies also demonstrated the release of nano-Ag from textiles [28-30]. Modelling

approaches have been used to estimate resulting NM quantities in different envi-

ronmental compartments, since these were not analytically accessible. However,

these approaches are limited to only a few NMs for which relevant life-cycle data

are available (for example data on production and release volumes, physical-

chemical data as well as information about environmental backgrounds) [25]. For-

Chapter 1

24

tunately, these include the most relevant types in terms of production volume and

use, such as TiO2 and CNTs. Table 1 summarizes the maximum predicted envi-

ronmental concentrations (max. PEC) for several (environmental) matrices,

based on a study that recently compiled this data from the available literature

[25]. Currently, it can be assumed that especially sediments may represent a sink

for unintentionally released TiO2 NMs and CNTs.

However, NMs can also be released to the environment intentionally. Application

of sewage sludge/biosolids that contain high concentrations of TiO2 NMs espe-

cially due to release from personal care products [31] to soils may increase the

exposure of this environmental compartment by a factor of 100 (Table 1.1).

Table 1.1: Maximum predicted environmental concentr ations (Max. PEC) of TiO 2 and CNTs

for different (environmental) matrices. Adapted fro m Gottschalk et al. [25].

Matrix Nanomaterial Max. PEC

Surface water TiO2 10 µg L-1

CNTs 10-3 µg L-1

Biosolids TiO2 103 µg g-1

CNTs 10-1 µg g-1

Sediments TiO2 104 µg kg-1

CNTs 103 µg kg-1

Soils TiO2 10 µg kg-1

CNTs 10-2 µg kg-1

Soils treated with sewage sludge TiO2 103 µg kg-1

CNTs 1 µg kg-1

Further intentional release occurs, if NMs are used in environmental remediation.

For example, iron nanoparticles (NPs) have been proposed for groundwater re-

mediation [32] and removal of chromium from wastewater [33]. The nano-form of

zero-valent iron (ZVI) is already considered an advancement over conventional

ZVI and is already applied in practice, while potential risks of nano-ZVI still re-

main unclear [34]. Apart from Iron-based NPs, also polymer-based NPs have


25

been tested for soil remediation purposes [35]. However, the most dramatic in-

crease in NM release can be expected, if NMs were used in agriculture, where

the development of approaches to utilize beneficial properties of NMs is increas-

ing since the 2000s. Amongst others, NMs are considered to be useful to en-

hance agricultural formulations, such as plant protection products or fertilizers

(e.g., see [36-40]). Such enhancements may be achieved for example by utilizing

the high surface area of NMs to reduce run-off losses of active ingredients or to

reduce release kinetics of nutrients. This aspect of nanotechnology is described

and analyzed in more detail in Chapter 2.

1.4 Environmental risk assessment of nanomaterials

Since 2000, there is a growing concern about the environmental safety of NMs [9,

41, 42]. Several adverse effects attributed to the presence of NMs have been

reported for microorganisms (e.g., [43, 44]), plants (e.g., [45, 46]), animals (e.g.,

[47, 48]) and humans (e.g., [49, 50]). The adverse effects are mostly attributed to

the increased surface area of NMs, which enables enhanced reactivity and/or

production of reactive oxygen species (ROS), which is the currently best de-

scribed indicator for NM toxicity [51]. In addition, NM surface properties, such as

coatings, surface treatments or surface excitation (e.g., by UV radiation) may in-

fluence effects of NMs on organisms [51]. This means that, even in aggregated or

agglomerated form, a NM can exert a (toxic) effect, for example by releasing a

toxic chemical compound from its surface.

Therefore, to prevent potential adverse effects arising from future NM applica-

tions in general their risk must be evaluated. In particular, this will be important

for NM applications into agricultural ecosystems which build the basis for human

nutrition. Risk (or hazard-) assessment is defined to be a product of “exposure”

and “effect” [52]. Thus, to thoroughly assess the “risk” of NM use in agriculture,

both the actual concentrations of NMs in the system under investigation as well

as their effects on the present organisms have to be known. Taking into account

the many material specific factors that may influence NM toxicity, it is evident that

no generalized risk assessment can be performed and thus that potential risks

Chapter 1

26

have to be evaluated for each NM separately. To address both exposure and ef-

fects, the project “Effects of NANOparticles on beneficial soil MIcrobes and

CROPS” (“NANOMICROPS”) was started within the Swiss National Research

Programme 64 (NRP64) [53]. This thesis constitutes the “exposure” part within

NANOMICROPS, consisting of (i) an assessment of current agricultural NM ap-

plications and thus the potential NM exposure (Chapter 2), (ii) the development of

analytical methods to assess the actual NM exposure in a model organism, im-

portant for ecosystem functioning (Chapter 3) as well as in agricultural soils

(Chapter 4 and 5) and (iii) description of NM fate in a model agricultural ecosys-

tem (Chapter 6). In a second thesis, conducted by Janine Moll, the biological ef-

fects of these NMs were evaluated [54].

1.5 Main nanomaterials investigated in NANOMICROPS

The main criteria for the selection of a NM to be evaluated within NANOMI-

CROPS were its general relevance as well as its potential use in agriculture, for

example as an active ingredient or additive in a pesticide or fertilizer formulation.

Both criteria were fulfilled by TiO2 NPs as well as CNTs, as they are among the

most produced and used NMs worldwide and are evaluated for use in agricultural

applications as well. Within this chapter the two selected materials are introduced

in a general manner, while in the second chapter their potential use in agriculture

is described and discussed in detail.

1.5.1 Titanium dioxide

Elemental titanium belongs to the group of transition metals. It constitutes approx.

0.44% of the earth’s crust, with concentrations in soils ranging from 0.1% to 0.9%

with a mean value of 0.35% [55]. There are three main crystal phases of TiO2:

rutile, anatase and brookite (Figure 1.2) [56]. In essence, all three structures are

based on the interconnection of TiO6 octahedra [56]. In the rutile and brookite

structure type, both edges and corners are shared while in anatase only edge-

sharing occurs (Figure 1.2) [56, 57]. Through their six corners/edges, the octahe-

drons are then connected to a 3-dimensional network. Rutile is the stable form,

whereas anatase and brookite are metastable and are readily transformed to ru-


27

tile when heated [56, 58]. The temperature range for conversion is between 400

and 1200°C, depending on the raw material and the processing methods. In addi-

tion, pure phases of brookite are difficult to obtain [58]. Therefore, currently only

rutile and anatase play a major role in industrial applications. However, new syn-

thesis routes for brookite have been established [57], which might open new ap-

plications for this crystal phase, e.g., for the use as an anode in lithium ion batter-

ies [57].

Figure 1.2: The three crystal phases of TiO 2, reproduced from [57].

Because of their high refractive index, their persistence and non-toxicity, rutile

and also partly anatase are among the most used white pigments worldwide [56].

TiO2 as a pigment is not only contained in products such as paints, plastics, pa-

pers, construction materials or ceramics, but also in food products [59] such as

chewing gums, cookies or sauces. The TiO2 material used in the latter is ap-

proved under the name E171, as “food-grade” TiO2 and may - to some extent -

contain nano-sized fractions [59, 60]. In any case, nanoscale TiO2 is the most

produced NM world-wide [23]. It is not only used as a pigment, but also as a pho-

tocatalyst [61, 62] with manifold applications, such as in air [63]/water [64] purifi-

cation, sterilization [65] and cancer therapy [66].

1.5.2 Carbon nanotubes

Carbon nanotubes are structures of sp2 hybridized carbon which are arranged as

rolled up graphene sheets (Figure 1.3). Diameters of such resulting tubular struc-

Chapter 1

28

tures are usually below 100 nm, while lengths of several micrometers can be

reached – qualifying CNTs as NMs according to the definition by the European

commission [8]. The first electron-microscopic evidence of such hollow nanome-

ter sized carbon filaments was given already in 1952 by the Russian scientists

Radushkevich and Lukyanovich [24, 67, 68]. However, due to the many re-

strictions in terms of exchange between the west and the east during the cold-

war, this knowledge lay quiet until the first experimental evidence of CNT synthe-

sis was published by Ijima in 1991 [69]. This first paper on CNTs with multiple

walls (Multi-Walled CNTs, MWCNTs) was complemented two years later by the

description of CNTs with a single wall (Single-Walled CNTs, SWCNTs) [70].

These works then were the driving force to initiate widespread CNT research

which increasingly explored the possibilities offered by this new type of material;

resulting in an increase in production capacity since 2006 by a factor of 10 to an

estimated value of 4.5 kilotons, as well as 24,000 annual scientific publications

and 3000 patents in 2011 alone [24].

Figure 1.3: (Single-walled)-Carbon nanotube structu re from a rolled up graphene sheet.

Modified from [71].

The main properties that are responsible for this enormous success of CNTs are

their low density, mechanical strength and their electrical as well as thermal con-

ductivity. Their combination of low density and remarkable mechanical strength -

tensile strengths of up to 300 GPa can be reached - makes them ideal building

blocks for lightweight but sturdy materials [72]. Their high electrical conductivity of

up to 5000 S cm-1 [73] and current carrying capacity of up to 109 A cm-2 [74] have

led to many applications in electronics and energy storage, such as CNT use in


29

transistors [75], in flexible thin film transistors for displays [76] or in lithium ion

batteries [77].

Another important property of CNTs that is also of environmental relevance is

their high hydrophobicity and connected sorption capacity. Hydrophobicity can be

of use e.g., in filters for water purification, allowing higher water fluxes and salt

rejection rates [78]. The high sorption capacity of CNTs can also be used to re-

move organic contaminants from waste water [79]. In the environment, however,

the high sorption capacity might lead for example to associations of CNTs with

organic micropollutants, which in turn result in locally elevated exposure concen-

trations of these micropollutants to organisms, enhancing their toxicity [80].

1.6 Challenges and analytical approaches for NM characterization

and quantification in complex matrices

1.6.1 General analytical challenges

In environmental matrices natural NMs are already present in large numbers [16]

compared to engineered NMs that - in addition - are often similar in their ele-

mental composition, size and shape. To make an iconic analogy, the task to de-

tect an engineered NM within a natural matrix such as soil can be regarded simi-

lar to the task of detecting a specific piece of straw in a haystack. Furthermore,

techniques that allow the observation and characterization of objects at the na-

noscale were initially tailored to the needs of mainly material scientists, who study

their objects of interest under clean conditions with a lot of a priori knowledge on

e.g., their elemental composition. Therefore, there are already a number of ana-

lytical techniques available to study NMs in very simple and clean matrices. With

regard to environmental matrices such as soils, sediments or sewage sludge,

these techniques however are not directly applicable to analyze NMs, as they

either lack specificity or are simply not per se compatible with complex samples.

This results in a current disparity between expected environmental NM concen-

trations (Table 1.1.) and actual detection limits that can be achieved, where the

Chapter 1

30

most specific analytical tools currently have a sensitivity in the milligram per kilo-

gram range at best [81].

Apart from detection and quantification alone, for some scientific questions (e.g.,

fate of the NM in the environment or toxicological mechanisms) additional

knowledge on some of their physical-chemical parameters within the studied en-

vironmental system may be required. For example, the uptake of Gold NMs into

cells was depending on size, shape and surface charge [82] and the toxicity of

CNTs to mice was highest when they resembled needle like and stiff structural

features rather than coils [48]. However, NMs may undergo changes in several of

these parameters while subjected to environmental conditions or under analysis.

As shown in Figure 1.4, NMs can interact with various constituents of the soil re-

sulting in changes of their physical-chemical parameters. Depending on the NM,

dissolution of the particles can take place over time, resulting in release of ions

that may exert toxic effects of their own. This can go along with changes in size

and shape or reactions with soil components that change the chemical composi-

tion or speciation of the NM. Furthermore, the NM may interact with soil compo-

nents such as dissolved organic matter that “coats” the particle and may modify

its colloidal stability and thereby its behavior in the soil pore water. These exam-

ples illustrate that the physical-chemical “character” of NMs in soil is subjected to

dynamic changes, leading to a discontinuity of NM properties [81].


31

Figure 1.4: Scheme, showing the different interacti ons, changes or modifications NMs may

undergo in soils or soil suspensions respectively, along with physical-chemical analytical

parameters that go along with these changes (red bo xes).

1.6.2 Sample preparation and storage

Considering the multitude of NMs and their individually different properties a

“one-size-fits-all” strategy cannot be developed and therefore only a few exam-

ples shall be given here.

If the goal is to describe the NM in its environment, the main challenge for sample

preparation is that it needs to be suitable for the analytical technique to be used

while preserving the NM properties to be analyzed. This can be very hard to

achieve, e.g., for NP size and agglomeration state in a complex liquid such as a

soil solution, as mentioned before (Figure 1.4). Especially the NP agglomeration

state is very sensitive to changes in the NP environment (pH, ionic strength etc.),

and changes in this environment necessary for the analytical technique may con-

sequently alter this property.

Thus, ideally sample preparation for a given NM containing sample is minimal.

Hyperspectral imaging microscopy (HSI-M) [83] is a promising technique that

may provide analysis of complex aqueous NM samples with minimal sample ma-

nipulation along with minimal time investment from sample collection to the actual

Chapter 1

32

analysis and is discussed and evaluated in further sections and chapters of this

thesis (see section 1.7.1 and Chapter 3).

If the goal is however, to identify and quantify a NM in a complex sample, it may

not necessarily be mandatory to preserve the structure of the sample. A selective

destruction of the surrounding matrix that leaves the NM intact or any other

means of separating NM and matrix from each other (e.g., separation by density

or charge) could be performed, as outlined in several case studies [81].

For sample storage, again dynamic processes such as agglomeration and sub-

sequent sedimentation have to be taken into account. Depending on the research

question, such processes will have to be mitigated (e.g., by sonication prior to the

measurement or by addition of stabilizing agents) or characterized and described

to the possible extent.

For inorganic particles with a high solubility, such as for example Ag NPs, disso-

lution must be taken into account, which affects the size of the NPs over time

[84]. For some NPs even the general experimental environment, such as temper-

ature or exposure of samples to sunlight may be of importance. Nanoparticles

with a coating sensitive to UV-radiation may agglomerate after photodegradation

of the coating [85], or ROS produced from nano-TiO2 may alter the NP environ-

ment (e.g., by oxidizing organic matter) and thus agglomeration status as well as

toxic effects.

In summary, some a priori knowledge on the NM of interest may be needed be-

fore the analysis, and the strategy then tailored to the material under investigation

and the scientific question.

1.6.3 Analytical strategies: The need for sophisticated analytical chains

The first steps of potential analytical chains are in principle comparable to con-

ventional strategies, as the analyte particles first have to be transferred into a

state suitable for the analytical technique to be used (e.g., a liquid or deposited

on a surface). This includes a separation of the particles from the matrix, a reduc-

tion of sample complexity, as well as steps that lead to an enrichment of the par-

ticles if necessary.


33

Thereafter, techniques specific for the different types of desired information can

be applied; for example, light scattering (size), ICP-MS (elemental composition),

electrophoretic mobility (surface charge), electron microscopy (shape) and X-ray

diffraction (crystal structure). Ideally such techniques are coupled on- or offline in

a sequence to simultaneously yield complementary information. This is especially

important, if the sample is not stable enough to be analyzed by different tech-

niques at different time points (e.g., on different days). Also, with the complexity

of NM containing environmental samples in mind, a methodological sequence

that is able to provide multiple lines of evidence for the presence of a specific NM

in a sample is of paramount importance.

In summary, different methods are required in combination to fully characterize

NMs in a given system. Such a –non exhaustive- methodological sequence shall

be briefly described in the following as an example for soils: (i) provided a specific

extraction technique exists, the NPs are transferred into a liquid dispersion, (ii)

the liquid dispersion is fractionated by size, (iii) the fractionation is verified by a

light scattering technique and the size is characterized, and (iv) the presence or

absence of the NPs of interest is proven by a specific measure of individual frac-

tions (e.g., elemental analysis in combination with an (electron microscopic) im-

aging technique). A similar approach is followed in chapter 4 of this thesis.

Note that in some cases, a single technology can be of use in multiple steps, for

example field flow fractionation can reduce sample complexity, but also give in-

formation about particle size. Single-particle ICP-MS has the potential to reveal

size distributions as well as the concentration of each size fraction. Electron mi-

croscopy visualizes NM sizes and morphologies, may quantify NPs in an auto-

mated fashion (see also chapter 5 of this thesis) and - coupled to elemental anal-

ysis such as EDX - also reveal the elemental composition of NMs.

1.6.4 Specific challenges and approaches for the analysis of TiO2 NPs in soils

Titanium is one of the most abundant elements in the earth’s crust. As mentioned

before, elemental concentrations in soils are therefore relatively high, ranging

from 0.1% to 0.9% with a mean value of 0.35% [55]. Titanium minerals are con-

sidered inert and therefore resistant to weathering. Similar to the problem de-

Chapter 1

34

scribed for CNTs, this great abundance and persistence of naturally occurring Ti

in soils results in the challenge of being able to distinguish between natural and

engineered (bulk or nano) Ti in this matrix. Consequently, there are very few re-

ports available aiming at identification of engineered Ti-NPs in environmental

samples [27, 86-88]. This identification however mainly relied on morphological

parameters that could be assessed by comparing the original particles with those

found in the sample by means of electron microscopy. This approach is also used

in Chapter 6 of this thesis.

1.6.5 Specific challenges and approaches for the analysis of CNTs in soils

Probably the biggest challenge for CNT quantification in soils is the high back-

ground carbon, present as inorganic carbon (i.e. carbonates) and organic carbon.

Organic carbon in soils can be classified according to its turnover time, i.e. stabil-

ity both against chemical as well as biological degradation, into labile (years),

intermediate (years to decades) and refractory (hundreds to thousands of years)

carbon [89]. A large fraction is constituted by labile to intermediate compounds,

such as humic acids – large organic structures that are formed after chemical and

biological decomposition of organismic residues -, other organic molecules (car-

bohydrates, proteins, organic acids etc.) and organismic debris [90]. The inorgan-

ic background in form of carbonates as well as the labile and intermediate organ-

ic background can be easily removed by acid fumigation [91] and thermal or wet-

chemical oxidation [92], respectively. These two methods have been combined to

a chemo-thermal oxidation method for the quantification of Black Carbon (BC)

(CTO-375, [92, 93]), which constitutes the remaining refractory organic carbon

fraction. BC is considered a continuum [94], which spans from partly charred

plant material over char and charcoal to graphite and soot particles with no clear

definition of boundaries between the different forms [95].

Sobek and Bucheli expanded CTO-375 to CNT quantification in soils [94]. CNTs

can be very resistant to such chemical and thermal treatments, depending on

properties such as their diameter and defects [94], and were thus found as well in

the refractory BC fraction with varying recoveries (26-93%). In addition to varying

recoveries, the ubiquitous presence of particularly soot in soils in the range of mg

per g hampers CNT quantification by such means. This problem holds true also


35

for other recently developed methods that make use of the CNTs thermal stabil-

ity, such as thermogravimetry coupled to mass spectrometry [96] or programmed

thermal analysis [97].

Another recent development related to thermal properties of CNTs is the detec-

tion and quantification of CNTs in complex matrices by microwave induced heat-

ing (MIH) [98]. The technique makes use of strong microwave absorption ob-

served for CNTs [99], resulting in intense heat release, light emission, outgassing

and even welding of CNT bundles. The very rapid temperature rise during micro-

wave irradiation is then used for quantification. This method is applied in Chapter

6 of this thesis to detect and quantify CNT uptake by clover plants.

Finally, spectroscopic methods have long been used for CNT characterization.

For example, CNTs produce specific signals in Raman spectroscopy [100] that

can be related to their structural properties. Consequently, Raman spectroscopy

in conjunction with microextraction in ionic liquids has been employed for

SWCNT detection in river water [101]. This method however has not been tested

to date in more complex matrices such as soils.

SWCNTs yield distinct near-infrared fluorescence emission when excited with

wavelengths between 600 and 800 nm. This feature has been used for their

quantification in water and sediments [102]. The method proved to be very sensi-

tive (in the order of ng L-1 and µg kg-1, respectively), however only working for the

single-walled CNT type.

The problem of co-isolation of CNTs and soot, as well as the lack of methods to

detect the multi-walled variants of CNTs specifically in soils were the starting

point for the developments presented in Chapter 4 of this thesis.

Chapter 1

36

1.7 Main analytical techniques employed in this thesis

1.7.1 Hyperspectral imaging microscopy

Hyperspectral imaging microscopy (HSI-M) is a recent advancement in conven-

tional light microscopy to both visualize and characterize NPs [83, 103] that has

been first introduced by the company Cytoviva®. The HSI part of the technique

was transferred to NP research from airborne remote sensing applications, where

it has been used and well established for several decades [104, 105]. HSI-M is

based on a combination of dark field illumination and visible to near infrared

(VNIR) spectroscopy. The NPs are illuminated in an enhanced dark field by a

collimated light source at oblique angles. Transmitted light is blocked by an aper-

ture and the NPs then appear as bright dots in front of a black background (Fig-

ure 1.5). Light that is scattered from these particles is then analyzed on a pixel by

pixel level by a VNIR spectrometer, resulting in a 3d-data cube that consists of a

2d image (e.g., a cell) with full VNIR spectral information in each pixel. The spec-

tral information of e.g., the analyte particles can then be extracted and joined into

a spectral library (SL). To then analyze a sample with an unknown composition, a

hyperspectral image (HSI) is acquired and a spectral mapping algorithm applied

that determines the similarity of the spectral profile in each pixel of the unknown

image and each spectrum in the SL of the analyte particles. Regions or pixels

containing the spectrum of interest are then mapped and highlighted in a color of

choice (Figure 1.5).


37

Figure 1.5: Schematic depiction of the HSI-M workin g principle. (A) Generation of a dark

field image of a NP from scattered light emitted at an oblique angle. (B) Extraction of spec-

tral information from a reference NP and (C) mappin g of the reference spectrum in a hy-

perspectral image of an unknown sample. Objects wit h a similar spectral profile as the

reference NP are highlighted in red.

The working principle of one of the most common mapping algorithms (spectral

angle mapping (SAM) algorithm) is displayed in Figure 1.6. SAM calculates the

similarity between the spectral profile in each pixel of the sample HSI and the SL

obtained from the reference HSI at each of the wavelength bands. This is done

by first determining a vector in n dimensions that represents the distance from the

origin (dark) to the light intensity recorded in each band of the sample spectrum.

The direction of this vector in n-dimensional space is used to define a unit vector

representing the sample spectrum. The same procedure is performed for the ref-

erence spectrum. SAM then determines the angle between the two unit vectors.

The best spectral match occurs when the angle between these vectors is the

smallest.

Chapter 1

38

Figure 1.6: Schematic representation of the spectra l angle mapping (SAM) algorithm used

to detect NPs in unknown samples based on reference spectral libraries. Taken from Cyto-

viva, hyperspectral user manual.

HSI-M possesses several advantages that make the technique very compelling

for NP analysis. Perhaps the most important one is the simplicity of sample prep-

aration. Many common techniques used for NP characterization easily introduce

artefacts through the necessary manipulation and sample preparation [106, 107].

This is because the behavior of NPs is very sensitive to changes in their envi-

ronment (e.g., pH, ionic strength, ligands that suppress or enhance agglomera-

tion) but also because particles may be altered by high energy radiation such as

X-rays [108] or electron beams [109]. In HSI-M, samples can be directly placed

on glass slides and analyzed without any further pre-treatment.

1.7.2 Asymmetric flow field-flow fractionation (aF4)

Field-Flow Fractionation (FFF) was first introduced by Giddings [110] in 1966. It is

a technique similar to liquid chromatography that enables the separation of mac-

romolecules, polymers and particles by size, making use of their diffusion against

an applied field instead of their interactions with a stationary phase inside a col-

umn. Although invented in the 60ies, commercialization of the technique began

not until the early 2000s by the two companies Postnova and Wyatt. Due to the

ease of online or offline coupling with a large number of other detectors or char-

acterization techniques (for a selection see Figure 1.7), numerous applications in


39

different disciplines emerged, spanning from protein [111] and polymer [112]

characterization to the measurement of environmental colloids [113] and NPs

[114] as well as engineered NPs [115-117]. In addition, several types of FFF

have been developed that make use of different separation fields (e.g., thermal,

electrical, hydrodynamic and gravitational). Among these, FFF with a hydrody-

namic field, i.e. flow-FFF, and especially its asymmetric variant (asymmetric flow

field-flow fractionation, aF4) has gained most popularity in the recent years. In

this thesis a commercial aF4 apparatus manufactured by the company Postnova

[118] was used, so the following descriptions refer to this system.

In aF4, the main difference to conventional liquid chromatography is that the sep-

aration takes place in a narrow channel consisting of an impermeable plate on

the top and an ultrafiltration membrane at the bottom, instead of a column (Figure

1.7 and 1.8). This setting is often advertised as being virtually free of a stationary

phase, thus minimizing particle losses that would otherwise occur due to interac-

tions with the column packing material. As this is essentially true per definition,

however, recent works [119, 120] - as well as this thesis - show evidence that,

depending on the NM and the experimental conditions, strong interactions be-

tween NMs and the membrane may take place, resulting in particle loss.

Figure 1.7: Schematic showing the simplified genera l setup of an asymmetric flow field-

flow apparatus and possible connected detectors. Mo dified from [121].

For particle separation, a hydrodynamic field is generated over the whole channel

length by a so called cross-flow pump (Figure 1.8). This pump actively sucks the

carrier liquid through the ultrafiltration membrane at the bottom of the channel

Chapter 1

40

(usually with a molecular weight cut-off of 10 kDa) which retains the particles. As

the top plate of the channel is impermeable an asymmetric drag field over the

channel length is created (Figure 1.8).

Figure 1.8: Generation of the asymmetric hydrodynam ic field in the channel of the used

aF4 apparatus. Modified from [121].

Before any particles can be separated, they first must be focused in a narrow

zone in the channel. This is achieved by applying a counter-acting flow (focus

flow) against the sample injection flow via a separate port (Figure 1.9). After this

focusing time – which should be selected according to the sample amount (mass,

volume) injected – the focusing flow is gradually turned off and the particles are

separated.


41

Figure 1.9: Sample injection and focusing procedure in the used apparatus. Modified from

[121].

Separation of the particles occurs due to their diffusion against the hydrodynamic

field (Figure 1.10) in a laminar flow profile. The particles form a diffusional cloud,

which extends towards the channel center, where the flow velocity of the carrier

is higher than at the channel bottom. For spherical particles and for a constant

hydrodynamic field force, the distance of each particle to the membrane interface

(accumulation wall) directly corresponds to its diffusion constant, according to the

Stokes-Einstein equation:

�� = �� (1.1)

where Dh is the diffusion coefficient, kB the Boltzmann constant, T the tempera-

ture, η the viscosity of the medium or carrier solution and rh the hydrodynamic

radius.

It is evident that, by relating to a specific diffusion constant, this equation also

directly relates to a corresponding rh. Thus, for particles with geometries different

from a sphere, rh will be an equivalent radius, relating to a sphere with the same

diffusivity. In addition, the Stokes-Einstein equation also shows that the smaller

the particles are, the higher their diffusion constant will be. Smaller particles will

therefore reach the region of higher flow velocity faster and will thus elute before

larger particles (“normal mode”, Figure 1.10). However, when the particle size

becomes larger compared to the diffusional cloud an inversion occurs, leading to

the faster elution of bigger particles (“steric mode”, Figure 1.10). As a rule of

Chapter 1

42

thumb, this – so called – steric inversion occurs at particle diameters above 500

nm.

Figure 1.10: Schematic showing the two different el ution modes in aF4. Modified from

[121].

The technique offers several advantages that will be applied in this thesis. The

first and certainly most obvious is the reduction of sample complexity and thereby

polydispersity, which is essential for the application of light scattering methods.

And second, it allows a physical separation of different size fractions that can be

further characterized. The latter point may be of special importance when it

comes to defining a material as a NM by means of number based concentrations.

A further advantage is the possibility of coupling aF4 to a series of detectors with

complementary information, which allows extracting a wealth of knowledge about

a sample: associations between particles and other solution constituents [122,

123], dispersion state, length [124] and shape [113, 125] of particles, particle

number concentrations [126] and more, depending on the used detector.


43

1.7.3 Multi-angle light scattering (MALS) detection

One such detector that can be coupled to an aF4 system is the multi-angle light

scattering (MALS) detector. MALS coupled to size exclusion chromatography has

been used since the 1990ies for the characterization of (branched) polymers

[127]. From MALS experiments, the molar mass and the size of polymers or par-

ticles are determined. In addition, MALS can also deliver information about con-

formation or shape of the analyzed polymer or particle, respectively. Due to these

possibilities, the application of MALS to CNT containing samples has been con-

sidered promising, as CNTs possess a relatively unique high aspect ratio shape

(see Chapter 4). The principles and theory of MALS have been described by Wy-

att in great detail [128-130]. Here, a short description of the working principle and

data evaluation will be given.

Basically, light scattering is an interaction of electromagnetic radiation with mat-

ter. When an electromagnetic wave illuminates an obstacle, be it a single elec-

tron, atom, molecule or particle, electrical charges (electrons and/or protons) are

set into oscillation by the electromagnetic field of the wave [131]. Thereby, these

charges become accelerated and scatter secondary electromagnetic energy in all

directions (Figure 1.11).

Figure 1.11: Simplified scheme of light scattering by an obstacle, according to [131].

Parts of the incident radiation are also transformed during this process into other

forms of energy (e.g., thermal energy). This transformation is called absorption

and is present in every scattering process [131]. As there are many different scat-

Chapter 1

44

tering processes, which to describe would be beyond the scope of this introduc-

tion and also of the MALS technique itself, the discussion in here shall be con-

fined to scattering by particles smaller than the incident radiation wavelength that

are highly diluted in an optically homogenous liquid, i.e. the region between the

Rayleigh and Lorenz-Mie scattering regime.

To do so, we can assume an arbitrary particle, which is conceptually divided into

regions. An incident wave of radiation (i.e. an oscillating field) induces a dipole

moment in each of these regions. The resulting dipoles scatter the incident radia-

tion in all directions. If the scattered radiation is observed at a particular distant

point, for a particle with a size very small compared to the wavelength, the sec-

ondary radiation is in phase and the intensity of the scattered radiation is homog-

enous for all angles [131]. As the particle size increases, the possibility for mutual

enhancement and cancellation of the secondary radiation increases (Figure 1.12)

[131]. This results in a scattering pattern of the intensity of the scattered light,

which is dependent on size, shape and angle of observation. Thus, observing the

result of the interaction of light with a particle, i.e. the modulated intensity of the

initial wave at multiple angles, gives information about the particle size as well as

its shape or conformation. Physical values derived from light scattering experi-

ments are absolute, meaning that they are calculated from fundamental physical

principles, with no need for calibration [132]. These principles will be introduced

in the following.

Figure 1.12: Light scattering by a large particle. 1-3-5 and 2-4-6 are paths of identical

length, whereas 1-3-7 is longer than 2-4-8. This de picts the differences in path length of

light scattered at different parts of a particle, w hich result in a reduction of scattering in-

tensity due to phase shifts. Figure reproduced from [132].


45

The basis for multi-angle light scattering measurements in dilute solutions or dis-

persions is given by equation 1.2 (Zimm equation) [132]:

(�)�� = ��(�) (1.2)

where θ denotes the angle between incident light beam and scattering direction

(θ=0°→straight forward, see also Figure 1.13), R(θ) the excess Rayleigh ratio, c

the concentration of the sample, K a contrast factor, M the molecular weight of

the sample, and P(θ) the particle scattering function.

Figure 1.13: The main components of a multi-angle l ight scattering measurement system.

The Rayleigh ratio is the angle dependent intensity of light scattered by a sample.

The additional term excess means that it is the contribution of molecules/particles

to the scattering of the entire solution, thus the difference between scattering in-

tensity of the solution and the pure solvent (eqn. 1.3). The light scattering intensi-

ty is thereby measured as a voltage obtained by photodiodes. Equation 1.3 de-

scribes the conversion of such a voltage to a Rayleigh ratio:

�(�) = (�(�)��(�,�� ))�!�"# = $ %(�)�%(�,�� )

%&'()* (1.3)

where I(θ) is the scattered light intensity of the solution, I(θ, solvent) the scat-

tered light intensity of the solvent, V the volume of the scattering solution and r

the distance between scattering volume and detector. The instrumental constant f

Chapter 1

46

is related to the apparatus geometry, structure and refractive index of the meas-

urement cuvette. U is a signal voltage. Division by the laser signal is performed to

compensate changes of the laser (power supply instability, temperature change,

laser aging). The constant f is usually determined once in a solvent with a high

Rayleigh ratio such as toluene, and stays constant for years, without the need for

frequent recalibration [132]. Besides the calibration of f, the detectors placed at

the various angles in the MALS system have to be normalized. As mentioned

before, very small particles scatter the laser beam in an equal intensity in all di-

rections. Thus, each photodiode at each angle should yield the same voltage,

when measuring such a particle. However, this is not the case in reality because

of the following reasons: differences in the build of the photodiodes result in dif-

ferent voltages at identical scattering intensities and each photodiode monitors a

different scattering volume due to the geometry of the setup [132]. Therefore, to

normalize the MALS detector, an isotropic scatterer is measured and each angle

is normalized to the detector at 90°, resulting in different normalization coeffi-

cients N(θ) for each angle. This coefficient is then integrated into eqn. 1.3, as

R(θ)=N(θ)f[(U(θ)-U(θ,solvent)/Ulaser]. The normalization coefficient is also usually

very stable, but is solvent dependent, and has thus to be adjusted if the solvent is

changed.

P(θ), the particle scattering function, denotes the change in scattered light inten-

sity in dependence on the angle of observation (θ). Thus it is generally defined as

the intensity of scattered light at angle θ to the intensity of scattered light at angle

0°:

�(�) = (�) (+°) (1.4)

The intensity of the scattered light declines with increasing angles due to phase

shifts of light beams scattered at different points of a large particle. For smaller

particles with a maximum size below λ/20 scattered light beams are in phase,

resulting in isotropic scattering (i.e. scattering intensities are constant over all ob-

served angles, Figure 1.14). Several mathematical expressions (not discussed

here for brevity) have been developed for P(θ) that are valid for certain defined

particle geometries or arbitrary shapes and where the so called radius of gyration

is included and can be extrapolated.


47

Figure 1.14: Variation of the scattering intensity in dependence of the angle of observation

for spherical polystyrene beads with a geometric ra dius of 10 and 50 nm (own measure-

ments).

The radius of gyration rg or – synonymously - the root mean square (RMS) radius

is the main value derived from a MALS experiment [132]:

-. = /∑12�2!∑12

(1.5)

with ri being the distance of the ith mass element of mass mi to the particles cen-

ter of mass. In other words, it is the sum of all possible radii to the particles gravi-

tational center. Hence, the rg contains structural information. The rg is a value en-

tirely derived from the light scattering measurement and is thus absolute and can

be determined without knowledge on concentration or refractive index increment

[132]. For irregularly shaped particles, such as branched polymers or CNTs, the

rg represents an average value, as these molecules or particles can have various

conformations where each of these possesses its own rg. This also highlights the

necessity to fractionate a sample of such particles before the light scattering ex-

periment into - close to - monodisperse slices, as for example provided by aF4

fractionation.

Chapter 1

48

1.8 Objectives and contents of the thesis within the NRP64 and

the project NANOMICROPS

The idea of using nanotechnology in agriculture and especially in plant protection

products and fertilizers is novel and under development since the early 2000s. As

such applications would increase NM flows into agricultural ecosystems dramati-

cally and negative effects of NMs on various organisms have been reported,

there is a need to proactively generate tools that allow a thorough risk assess-

ment of such NMs.

Overall, this thesis therefore aimed to contribute:

- to the knowledge on the current state of developments of nanotechnology

in plant protection and fertilization and therefore on the potential exposure

of agricultural ecosystems to NMs in the future by assessing the current

situation regarding development and use of NMs in plant protection prod-

ucts and fertilizers.

- to novel analytical methods that allow to describe the actual exposure of

agricultural ecosystems to NMs in the future.

- to the initial knowledge on the fate of such NMs in a mesocosm study with

natural soil.

The chapters of this thesis are organized as follows:

Chapter 2 contains a critical assessment of the current state in research and in-

dustry regarding the development and use of NMs in plant protection and fertiliza-

tion. The focus was set on these specific two applications as they would pose the

most significant source of intentional NM input into agricultural ecosystems. The

historical development of the field was analysed as well as types of materials and

forms in which they appear in agricultural formulations. In addition, research

needs are highlighted and the current state regarding the legislation is discussed.

Chapter 3 describes how HSI-M could be used to quantify NMs in a complex

model system, namely cell cultures of the protozoan Tetrahymena thermophila.

Protozoa play an important role in aqueous ecosystems but also in soils, where

they take key roles in the carbon and nitrogen cycle [133], making them invalua-


49

ble for ecosystem functioning. Thus, investigating NM uptake behavior of these

organisms is of great interest. Analytically, the influence of extracellular sub-

stances as matrix constituents is evaluated as well as the specificity of the spec-

tra acquired from reference samples. Parameters that influence this specificity

are highlighted. Finally, semiquantification of NM uptake into protozoan cells was

performed over two time points using HSI-M and the results are discussed.

Chapter 4 describes a novel method to detect CNTs in soot and soil samples

based on shape. The method makes use of aF4 to reduce sample complexity

and MALS to describe the analytes in terms of shape. The method was devel-

oped in increasingly complex matrices, i.e. from suspensions of pure materials

(CNTs and soot) to real soil samples. Different method validation and quality con-

trol measures were taken and are evaluated. Ultimately, the usefulness of the

method with regard to possible analytical chains as well as predicted environ-

mental concentrations is critically discussed.

In chapter 5, the preliminary results of an attempt to obtain number based con-

centrations of pure CNT suspensions from automated electron microscopy image

analysis are presented. In addition, first images of CNTs extracted from soil sus-

pensions are shown.

In chapter 6, the fate of TiO2 NPs and CNTs is monitored in a model mesocosm

experiment with clover and wheat plants, where biological as wells ecosystem

effects of the employed NMs were assessed. The method for CNT detection de-

scribed in chapter 4 of this thesis is employed to detect CNTs in different soil

depths. In addition, the novel microwave induced heating method [98] is applied

to detect and quantify CNT fractions in the plants. This experiment has been

conducted within the frame work of NANOMICROPS and the NRP64 in collabo-

ration with Janine Moll.

The last chapter summarizes the findings of the previous chapters and points to

additional methodological improvements as well as experimental approaches re-

garding various aspects of TiO2 NP and CNT analysis in environmental samples.

Chapter 1

50

51

Nanomaterials in plant pro-Chapter 2

tection and fertilization: current state,

foreseen applications and research pri-

orities

Alexander Gogos, Katja Knauer and Thomas D. Bucheli

Reprinted with permission from: Journal Of Agricultural And Food Chemistry

2012, 60, 9781-9792. Copyright 2012 American Chemical Society.

Chapter 2

52

Abstract

Scientific publications and patents on nanomaterials (NMs) used in plant protec-

tion or fertilizer products have exponentially increased since the millennium shift.

While the United States and Germany have published the highest number of pa-

tents, Asian countries released most scientific articles. About 40% of all contribu-

tions deal with carbon-based NMs, followed by titanium dioxide, silver, silica, and

alumina. Nanomaterials come in many diverse forms (surprisingly often >> 100

nm), from solid doped particles to (often non-persistent) polymer and oil-water

based structures. Nanomaterials serve equally as additives (mostly for controlled

release), or active constituents. Product efficiencies possibly increased by NMs

should be balanced against enhanced environmental NM input fluxes. The dy-

namic development in research and its considerable public perception are in con-

trast with the currently still very small number of NM-containing products on the

market. Nano-risk assessment and -legislation are largely at their infancies.

Nanomaterials in plant protection and fertilization: current state, foreseen applications and research priorities

53

2.1 Introduction

Materials, whether of natural or manufactured origin that possess one or more

external dimension in the range of 1-100 nm are defined as nanomaterials (NMs)

[8] and are increasingly used in a wide range of technical applications and con-

sumer products due to the unique physicochemical properties emerging at the

nano-scale. Key applications of NMs can already be found for example in the ar-

eas of electronics, energy, textiles, pharmaceutics, cosmetics and biomedicine.

This trend is fueled by many hopes and promises, such as improved performance

and new functionalities accompanied by a significant reduction in the use of re-

sources and the generation of waste. Thus, NMs are generally believed to in-

crease profitability and sustainability [134].

As a constantly growing world population is in demand for higher agricultural

yields, and as important resources such as phosphorous and potassium are lim-

ited, more effective strategies to optimize agricultural practices are urgently

needed. Hence, attempts to do so with the help of nanotechnology and NMs are

increasing [36-38, 40, 135-137]. The application of NMs in agriculture aims to

reduce applied amounts of plant protection products (PPP), minimize nutrient

losses in fertilization, and increase yields through an optimized nutrient manage-

ment [36-40, 137-139]. Several factors that influence the efficiency of PPP and

fertilizers could be addressed using NMs. For example, active substances can be

lost during application through drift, run-off, evaporation, photolysis and hydroly-

sis and degradation by microorganisms. As additives, NMs with a high surface

area and appropriate sorption properties may minimize losses by reducing run-off

and decreasing release kinetics. Specifically designed particles could furthermore

protect active ingredients from photodegradation or enhance uptake into the

leaves and other parts of the plant. Nanomaterials may also substitute hazardous

organic (co-)solvents, present in some PPP, and facilitate their dispersion, e.g.,

on plant surfaces. As active ingredients, NMs could reduce application rates

through their enhanced reactivity.

Despite these expected positive impacts in various fields, some NMs definitely

have properties that classify them as potentially hazardous [140]. Therefore, a lot

of attention is currently paid on the potential risks arising from these materials

Chapter 2

54

[141], which has already led to a number of studies that examine their mecha-

nisms of unintentional emission and toxicity [26, 142, 143].

However, the use of NMs in agriculture – and specifically in plant protection and

fertilization – may pose unforeseeable risks, in particular because these applica-

tions comprise an intentional input of NMs to the environment. This may lead to

higher input fluxes than predicted to date (Table 2.1) [144]. Consequently, human

and environmental exposure due to NM residues in crops and soil might increase

accordingly, with exposure routes including possible bioaccumulation of NMs in

the environment [145, 146], and in the food chain [147, 148]. Agriculture is aiming

at a sustainable management of natural resources which is globally imperatively

needed as set out by the EU Millennium Development Goals [149]. Therefore, the

application of NMs must be critically evaluated to guarantee their safe use for

agriculture.

With nanotechnology related markets growing at an enormous speed, there is an

urgent need to regulate products with nano content. This need for regulation is,

however, in general – but also specifically for agriculture – adversely accompa-

nied by a lack of knowledge on the current state. In addition, the unique proper-

ties of NMs alone already pose many difficulties to the regulatory bodies, begin-

ning with the statutory definition of a NM [150]. Regulatory issues in this context

were up to now only reviewed regarding NMs in food production in general [151].

Agricultural NM applications and their potential advantages are often mentioned

in the literature. Although several authors briefly addressed diverse prospective

agricultural NM applications, without being specifically focused and comprehen-

sive on PPP and fertilizers [36, 39, 40] (for a compilation of reviews related to the

field, see Table S2.1), a systematic compilation of NMs in plant protection and

fertilization, including a list of existing patents, is currently lacking.

In this review, we therefore like to align the increasingly common perception of

NM use and benefit for crop protection and cultivation with actual scientific facts

and figures. To achieve this goal, we will first give an overview on the develop-

ments concerning nano-PPP and fertilizers since the beginning of the 21st centu-

ry and of the current state in the Nano-Agro-Business. In the following, we will

systematically discuss different nano-materials and their properties that are envi-


55

sioned to improve agricultural formulations, based on the scientific literature and

published patents. To conclude, we identify specific research gaps related to the

risk assessment of NMs in agriculture, and address relevant aspects of nano-

legislation.

2.2 Methodology

For this compilation, the following databases and literature sources were used: (i)

scientific databases: Web of Knowledge, Google Scholar, (ii) Patent databases:

World Intellectual Property Organization (WIPO, EU), Free Patents Online

(EU+US), Web of Knowledge, and (iii) grey literature obtained from various

sources. Literature used for this evaluation was selected when it referred specifi-

cally to the development, testing and application of nano-plant protection and fer-

tilizer products. Remediation applications were excluded, as they were out of the

scope of this review. In total, we selected a total of 36 publications and 33 pa-

tents (listed systematically in Table S2.2 & S2.3) that were published until begin-

ning of February 2012. In the following chapters, these 69 articles will be looked

at from different angles.

Furthermore, concerning input of NMs into soils, we calculated application rates

and fluxes resulting from such applications (Table 2.1). For the calculations, we

assumed an application volume of 300 L ha-1, 20 cm plow depth, an average soil

bulk density of 1.4 g cm-3 , and one application per year.

Chapter 2

56

Table 2.1: Modeled fluxes 1 of different NM s and application rates of PPP or fertilizers,

selected from scientific literature and patents.

NM Modeled flux into soil Ref. Application rate and

calculated flux from

PPP/Fertilizer 2

Ref. Flux r a-

tio 3

TiO2 Realistic scenario:

0.4 µg kg-1 y-1

[153]4 4.5-15 kg ha-1 ≈

1607-5357 µg kg -1 y-1

[154] 334-1116

High exposure scenario:

4.8 µg kg-1 y-1

7.5 g ha-1 ≈

2.7 µg kg -1 y-1

[155] 0.56

0.28-1.28 µg kg-1 y-1

(US, EU and CH)

[156]5 Max. 30 kg ha-1

≈ 10714 µg kg -1 y-1

[157] 2232

Ag NP Realistic scenario:

0.02 µg kg-1 y-1

[153] 4 15 g ha-1 ≈

5.4 µg kg -1 y-1

[158] 54


0.1 µg kg-1 y-1

8.3-22.7 ng kg-1 y-1

(US, EU and CH)

[156]5

CNT Realistic scenario:

0.01 µg kg-1 y-1

[153] 4 3-12 g ha-1 ≈

1.1-4.3 µg kg -1 y-1

[159] 55-215


0.02 µg kg-1 y-1

0.56-1.92 ng kg-1 y-1

(US, EU and CH)

[156]5

1 Limited to those NMs for which data due to usage in the anthroposphere was available.

2 Assuming an application volume of 300 L ha-1, 20 cm plow depth, a soil bulk density of

1.4 g cm-3 [152] and an application once per year. 3 Calculated as Flux from PPP/Fertilizer divided by the value of the highest modeled flux. 4 Based on an annual substance flow analysis from products to soil in Switzerland.


57

5 Based on a probabilistic material flow analysis from a life cycle perspective of engineered NM

containing products.

2.3 Evolution of research and development activities for NMs in

agriculture

Historically, the notion that NMs could be of use in agricultural systems is a fairly

new one and under development now for approximately a dozen years (Figure

2.1). Apparently, the US National Nanotechnology Initiative (US NNI) in the early

2000s [160] coincided with the onset of this new technology. The US NNI also

invested in possible applications of NMs in agriculture by supporting such re-

search in the US department of agriculture (USDA), yet in 2005 with 0.5% of the

total NNI funds only, [161] and remaining at this level till 2012 [162]. The Europe-

an Union (EU) funded nanoresearch as well since the beginning of this millenni-

um with the FP6 (2002-2006) and FP7 (since 2008), however with no specific link

to agriculture. In the field of NMs and agriculture, the annual output of NM-related

scientific publications is still drastically lower than in other fields: from 2000 to

2011 a general search using “(nanoparticle* or nanomaterial*) and (agricult* or

agronom*)” yields 483 hits in web of knowledge, whereas “(nanoparticle* or na-

nomaterial*) with the category refinement “material science and engineering”

yields 275.338 hits (search performed on 16.5.2012). Furthermore, there is a

huge discrepancy between the still relatively small amount of published peer re-

viewed papers and patents on agricultural NMs, and the public discussion and

grey literature. Lagging behind scientific publications approximately 2-3 years, an

exponential increase in open-source publication activity in this field started

around 2006 (Figure 2.1). Google delivers 786 hits on the term “nanomaterials

and agriculture” in the year 2000. In 2011, hits reached a number of 281.000. In

more detail, the activities concerning “nanomaterials and pesticides” also in-

creased exponentially (Figure 2.1), reaching 188.000 hits in 2011. Publications in

the field of NMs and fertilizers, however, are less prominently discussed, with

77.100 hits in 2011 (searches performed on 20.11.2011).

Chapter 2

58

Figure 2.1: Temporal development of publication act ivity referring to NMs in agriculture

since 1990: as a rough search in the web of knowled ge (WoK, right axis, blue squares), as

publications and patents specifically selected (rig ht axis, red dots) and searches on

google (left axis, empty symbols). Searches were pe rformed on 20.11.2011.

In 2004, the article “Down on the Farm” was released by the Action Group on

Erosion, Technology and Concentration (ETC group) [161], which may constitute

a turning point for the topic in public discussion and has been cited rather fre-

quently (search on google: “down on the farm ETC group”: 1250 Hits, search per-

formed on 18.11.2011). It tackled a diverse list of topics, spanning from nano-

encapsulated pesticides to nano-sensors for pathogen detection and many more.

Yet on the NM-pesticide topic, the focus was clearly on micro (or nano)-

encapsulated pesticides, referring to already commercialized products of Syngen-

ta (Primo/Banner MAXX with diameters down to 100 nm and ZEON) and some

patents by BASF [163] and others. The main statements in this article were that

NM applications in food production and agricultural industry are just emerging –

therefore being overlooked even by nano-technologists – but may exceed the

impacts of farm mechanization and the “green revolution”. In 2005, an article

concerning “applications of knowledge in development” was released by the Task

Force on Science, Technology, and Innovation within the UN-Millennium Project


59

[164]. This article stated that nanotechnology in agriculture would be especially

interesting to developing countries, bearing the potential to reduce hunger, mal-

nutrition and child mortality. Thus, it may not be a coincidence that emerging

economies with a large agricultural sector and population, such as China and

India, are having a greater interest in using NMs in agriculture as demonstrated

by a relatively high output of publications (Figure 2.2). In this article, China, India

and Korea are declared as the frontrunners in nanotechnology in the developing

world, which seems to be reflected in Figure 2.2. Looking at institutions that con-

tributed to publications in the field, it is surprising that both agriculture and nano-

specific research institutes contributed only approximately 17 % to the publica-

tions, whereas the majority originates from chemistry departments (19%). How-

ever, when it comes to the number of patented technologies, the leading coun-

tries are those of the Western World such as the United States and Germany.

Most patents are held by companies (56%) that are mainly represented by small-

er enterprises (Figure S2.1). However, BASF is holding 27% of all company pa-

tents and 15% of the total. Universities (29%) and individuals (15%) share the

remainder. These are some indications that there is an increasing endeavor and

progress, especially from industry, in the development of agricultural nano-

formulations in the last years (Figure 2.1, also see Figure S2.1 and Table S2.3).

This is, however, not yet reflected by data on commercialized products (Table

S2.4), which is still very scarcely available and often not conclusive, regarding

their actual NM content and speciation. Actually, almost no “nano” PPP seem to

be currently on the market, unless there are some that are not declared.

Chapter 2

60

Figure 2.2: Countries active in research and develo pment of nano-plant protection and

fertilizer applications.

2.4 Classes of NMs and their intended purpose in agriculture

Generally, the elementary composition of NMs in PPP or fertilizers (Figure 2.3)

can be based on carbon (i.e., carbon nanotubes (CNT), liposomes, organic poly-

mers etc.), metals or metal oxides (i.e., silver (Ag), zinc oxide (ZnO)), metalloids

(silica), and nonmetals (sulfur (S)). The frequency of their application followed the

order: lipids/polymers/emulsions > titanium dioxide (TiO2) > silver> silica (Figure

2.3). Note that the order changed to lipids/polymers/emulsions > TiO2 > ZnO and

copper oxide (CuO), taking into account the patents only. Nanomaterials can be

present in formulations as solid particles or as non-solid structures (Figure 2.4).

The latter can be lipid or polymer (natural or synthetic) based structures or oil-

water (O/W) emulsions. The size of such structures in both patented and pub-

lished formulations varied mostly between 100-300 and 300-2000 nm (Figure

S2.2). In patents, this size range was even extended over 2 µm in a few cases.

Size fractions below 100 nm could be found in 37% of the patents and 54% of the

publications. This fraction is lower, however, if provided nominal sizes (size not

confirmed in the actual formulation) were excluded. Hence, a considerable frac-

tion of the formulations indicated to be ”nano” do contain size fractions extending


61

beyond the “nano-range” (i.e, >100 nm; [8]). With 74% of all papers, PPP pre-

vailed over fertilizer products (Figure 2.5A). Overall, 35% of the PPP were

planned to be used as fungicides, 33% as insecticides and 27% even were con-

sidered for multiple categories (Figure 2.5B). In 41% of the PPP, NMs were the

active constituent (Figure 2.5C). This was especially the case for Ag and S NM

applications, where the NM itself was used as a fungicide [158, 165-172]. Further

examples can be found in the case of silica [173, 174] and alumina [175], which

were tested as insecticides. Another example is the BASF patent

WO/2011/067186 [176] based on nano-copper salts acting as a fungicide. How-

ever, in 57% of the PPP, NMs were additives (Figure 2.5C) that acted as con-

trolled release carriers (Figure 2.5D, 56%), protective (15%) or dispersing agents

(11%) and photocatalysts (11%). Apart from this, also milling of conventional

crystalline solid pesticides to a sub-micron range was performed, improving bio-

logical efficiency [177]. Alternatively, emulsions can be spray dried to nano-

powders and then be re-dispersed [178]. With the exception of nano-S, which

was covered by one publication only [172], in the following, individual NMs will be

discussed in the order of their frequency of application (Figure 2.3).

Figure 2.3: Number of different NMs used in plant p rotection and fertilizer publications and

patents.

Chapter 2

62

2.4.1 Non-solid NMs

Today’s discussion on NMs in general mostly refers to NMs of a solid particulate

nature. In agriculture (and maybe elsewhere as well), however, the most promi-

nent fraction of NMs is non-solid, comprising nanoscale structures that may for

example encapsulate an active ingredient in a PPP (Figure 2.3, 2.4). Active sub-

stances are often poorly soluble in water and at room temperature even solid or

crystalline, and therefore brought into solution with organic (co-)solvents. To

avoid the use of the latter, one method of choice is the use of O/W-based emul-

sions [157, 179-182] (Figure 2.4) that enhance the solubility and thereby the load-

ing capacity of the formulation for the active substance. Furthermore, this may

also enhance coverage of the hydrophobic leaf surface and penetration of the

active substance through the cuticula. However, nano-emulsions are meta-stable

systems, which are prone to crystallization, agglomeration and sedimentation.

Stabilization is often achieved by sufficient amounts of suitable surfactants and

additional protective colloids [183]. One example of a successful application of

this technique is the BASF Patent WO2011138701 [184]. Therein monoglycer-

ides, i.e., hydrolyzation products from natural fats, are used as an amphiphile to

form liquid crystal- and microemulsion-structures (Figure 2.4) that are able to in-

corporate up to 30-40% of water in the oil phase. The elements of such nano-

structured O/W emulsions are thought to act as if they were the bulk substance.

Structures created this way are quiet labile, as for example addition of ethanol

can instantly destroy them. Furthermore, nanostructured monoglycerides also

occur naturally in the gastro-intestinal tract, after action of the pancreatic lipase

enzyme that hydrolyses triglycerides to monoglycerides [185].


63

Figure 2.4: Physical appearance of NMs in the analy zed publications and patents, classifi-

cation adapted and adjusted from [38, 151, 186].

It is also possible to create amorphous solid organic NMs by spray drying an O/W

emulsion into a redispersable powder, as shown by Elek et al. [178]. After redis-

persion, they found amorphous nanoparticles of Novaluron in an O/W emulsion.

The nano-formulation of novaluron could, however, not surpass the activity of a

commercial non-nano emulsion.

Examples of polymers used include nanospheres of polybutylcyanoacrylate [187],

polyethylenglycol [188] or polyvinylpyrrolidon [189]. Such materials are often

used because they are already known from medical applications. However, sub-

stances approved for medical applications may involve higher risks than such

approved for use in food.

Other non-solid structures that can be used in this context are liposomes (Figure

2.4). Liposomes are spherical bilayer vesicles formed by dispersion of polar lipids

in aqueous solvents. Liposomes have similarities to the structure of biological

membranes, hence they are highly biocompatible and therefore usually biode-

Chapter 2

64

gradable [190, 191]. A special form of lipid based NMs are cochleates (Figure

2.4), in which a lipid layer sheet is rolled up in a spiral fashion [192]. Another ex-

ample for promising biocompatible substances considered for agricultural use are

chitin derivatives [193, 194]. Chitin is the most abundant natural amino polysac-

charide [195] and is, together with its derivative chitosan, a very prospective mol-

ecule for encapsulation techniques. These nanocapsules (Figure 2.4) have very

interesting possible applications, for example in cosmetics, food and nutrition,

pollutant capture from wastewater, drug delivery and many more [195] and there-

fore seem worth further research. From an environmental point of view, they

seem of no concern. Chitosan is a major component of the cell walls of common

soil fungi and is produced by deacetylation of chitin. Chitosanase is known since

the mid seventies to be widespread among microbes [196].


65

Figure 2.5: Purposes of agricultural NM application s (A), types of PPP containing NMs (B),

general functions of NMs in PPP (C), and tasks of N M additives in PPP (D)

A company that attempts commercialization of such NMs is Vive Crop (formerly

Vive Nano Inc., [194]). With an actual budget of 16 million USD, they cooperate

with major companies, incorporating their active ingredients into chitosan or poly-

acrylic acid nano-formulations. However, until now, they state that none of their

products has hit the market, but they are optimistic to reach a breakthrough in

2013 [197]. One of the few examples of already commercialized formulations in

this context is Banner MAXX® [198] of Syngenta, belonging to the class of emul-

Chapter 2

66

sions. In “Down on the farm” it was stated that this formulation contains “extreme-

ly small particles of about 100 nm”. Again, concerning the abovementioned struc-

tural specialty of emulsion-nanostructures, it may be misleading to use the term

“particles” in this context. Moreover, Banner MAXX® uses a biodegradable agent

(tetrahydrofurfuryl alcohol) to dissolve the active ingredient.

2.4.2 Solid NMs

2.4.2.1 Titanium dioxide (TiO2)

Since its commercial production in the beginning of the twentieth century, TiO2 in

its bulk form has been widely used as a pigment, especially in paints [199]. In this

context it is already used in agriculture to exert a marker effect during spraying of

agricultural chemicals in amounts of 10 to 25 % of weight [200]. In 1972, Fujishi-

ma and Honda discovered the phenomenon of photocatalytic splitting of water on

a TiO2 electrode under ultraviolet light [201]. Since then, a lot of research has

been done to explore the potential of this material in its nano form especially in

the categories “energy” and “environment”. Concerning the latter, the main prop-

erty of TiO2 that has been exploited is its photocatalytic activity. It is generally

regarded as a highly efficient environmentally benign photocatalyst [199]. Main

applications therefore also include degradation of pesticides or pollutants in soil

remediation [202, 203].

However, one disadvantage of TiO2 NMs is that they are mostly active in the

presence of UV-light, due to their large band gap of approximately 3.2 eV [204].

This has implications for their use in agriculture, because the majority of sunlight

consists of visible light and only to ~5% of UV light [204]. Modifying TiO2 NMs

with different metals or other elements is a widely used technique to alleviate this

problem and enhance photocatalytic activity by shifting the band gap response of

TiO2 NMs to the visible region [199, 204]. Materials that are used for this purpose

are for example semiconductors with a more narrow band, metal or non-metal

ions and others [204]. In this context, a bactericidal activity of TiO2 due to light-

induced oxidizing reactions has been shown in Escherichia coli [205] and has in

that respect also been investigated for use in fungicide applications [206]. Lu et

al. [206] reported the use of Cerium (Ce) -doped TiO2 NM as an active substance

to control the downy blight disease occurring on litchi plants (Litchi chinensis) and


67

powdery mildew on cucumber (Cucumis sativus L.). In lab experiments, they

could show that the antifungal effectiveness of their NM was dependent on the

available light source, decreasing in the order black light (315-380 nm)>sun

light>indoor light. The results of their field experiments showed that the effective-

ness of 1.0% Ce3+- doped TiO2 was able to exceed that of the commercially

available Degussa P25.

In addition, the photocatalytic activity of (modified) TiO2 NMs can be used to re-

duce half-lives of pesticides, while ideally maintaining their effectiveness. This

could be demonstrated by Guan et al., who reported two types of photodegrada-

ble insecticides [155, 207]. In both cases, TiO2 was used as a photocatalyst to

enhance degradation of the used pesticide Imidacloprid and Avermectin respec-

tively. The effectiveness of the nano-Imidacloprid formulation was then tested

against adults of a storage pest beetle (Martianus dermestoides) and found to be

more effective than the conventional non-nano Imidacloprid (LC50 9.86 vs. 13.45

at an application rate of 25 mg L-1). Compared to other formulations, this applica-

tion would cause the lowest of all calculated TiO2 fluxes into soil (Table 2.1).

Similar results could be obtained using tungsten-doped TiO2 NMs in the Aver-

mectin formulation. Following up on the Imidacloprid-study, the formulation was

tested in the field [208], where shorter Imidacloprid half lives in soil were ob-

served for the nano-Imidacloprid formulation compared to the non-nano-

Imidacloprid.

However, TiO2 NMs can also be used in the opposite way, protecting a system

from photodegradation. This is achieved by employing the less photocatalytically

active rutile as the crystal form and coating the particles with different functional

moieties. Many active substances in PPP are sensitive to sunlight and therefore

prone to photocatalytic degradation. Shielding the active substance from radiation

is therefore believed to lower application rates of the PPP. An example for this is

the BASF patent WO/2009/153231 [154]. As a coating, an aluminum-oxide

(Al2O3) layer linked to a silicon containing polymer (Dimethicone) is proposed.

This particle, commonly known as TiO2-M262 (Sachtleben Chemie/Kemira) is

also used in sunscreens. The protected active substance was Metaflumizone at

10 g ha-1. The residual activity of the PPP was monitored for up to 10 days and

found to be 21% (Controls), 57% (common polymeric UV-protection agent) and

Chapter 2

68

64% (the latter mixed with the coated nano-TiO2). This would mean an increase

in efficiency due to the addition of TiO2 NMs of only 7%. In the experiments de-

scribed in the patent, TiO2 concentrations of 15 to 50 g L-1 with application rates

of 300 L ha-1 were employed. Thus, 4.5 to 15 kg of titanium dioxide would be ap-

plied to 1 ha. This would imply a flux of 1607 to 5357 µg kg-1 per application,

which is four thousand fold higher, than the annual flux estimated by Gottschalk

et al. [156] (Table 2.1).

Moreover, not only active substances in PPP are worth being shielded: Evonik-

Degussa developed a “superspreading” agent containing TiO2 as a photoprotec-

tive constituent to shield plant leaf surfaces from UV-light [209]. This treatment

then is intending to reduce sunburn damage in the plants and thereby loss of

yield.

Apart from light based modes of action, TiO2 can also be used as a dispersing

agent, as demonstrated in a patent by Rhône-Poulenc [157]. Therein, up to 100 g

L-1 TiO2 in a size range between 100 nm and 1 µm is employed as a fine powder

with a hydrophilic surface. Further described advantages of the use of TiO2 NMs

in agricultural products are the enhanced growth of spinach after incubation with

nano-anatase TiO2, reported in several works [210-214]. The improved growth

was related to a reduction of N2 to NH3 in the spinach leaves.

2.4.2.2 Silver (Ag) NMs

Unique optical and physical properties have led to the use of Ag NMs in catalysis,

construction of highly sensitive and selective detectors, optical (bio-)labeling,

conductivity elements in electronics, sensing, and many more [215, 216]. The

antibacterial properties of Ag and even nano-Ag, however, have been known for

centuries [18, 217]. Products containing nanoscale Ag particles have been com-

mercially available for over 100 years and were used in applications as diverse

as pigments, photographics, wound treatment, conductive/antistatic composites,

catalysts, and as a biocide [18]. Several studies revealed that Ag NMs may ex-

hibit the same antimicrobial properties associated with ionic Ag, which lead to an

increased use of Ag NMs for disinfection purposes in a wide spectrum of con-

sumer products such as fabrics, plastics, papers, and washing machines [216].

Above all, Ag ions are believed to have a low toxicity to animal cells [217].


69

Concerning PPP and fertilizer applications, Ag NMs are solely investigated as

fungicides. For example, Jo et al. [165] tested Ag NMs against two plant-

pathogenic fungi, Bipolaris sorokiniana, which infests important agricultural crops

such as Hordeum vulgare and Zea mays, and Magnaporthe grisea, a rice-

pathogen. They discriminated between Ag Ions and Ag NMs by comparing Ag-

NO3 to Ag NMs. The determined EC50 values for the different applications were

up to a factor of five lower for AgNO3 than for Ag NMs regardless of the pathogen

species. The effective concentrations of Ag NMs were approximately in the same

range as in the study of Kim et al. (2008) [158]. They tested the effectiveness of

Ag NMs against powdery mildew occurring on roses (Sphaerotheca pannosa var.

rosae) in the field [158] and observed a decline in mildew infestation of 95% a

week after the application of 15 g Ag NM ha-1. According to ref [218] application

rates of conventional fungicides against powdery mildew range from 105 g to 6

kg per ha. This shows that compared to other conventional fungicides, Ag NMs

could be applied in lower amounts, achieving the same effect. However, the flux

arising from such an Ag NM application is still 54-fold higher, than modeled for a

high exposure scenario (Table 2.1), and even 245-fold higher than the maximum

flux per year, as modeled by [156].

Alavi and Dehpour [168] could also show a higher efficiency of a commercial Ag

NM-product (Nanocid L2000) compared to a conventional fungicide. However, Ag

– in concentrations occurring in sewage sludge – has the potential to severely

affect microbial biomass in soils [219]. Silver NMs are releasing Ag ions in the

process of aging [220], thus it is generally questionable from an environmental

point of view to introduce even more particles (Table 2.1), and consequently Ag

ions, into soils.

2.4.2.3 Silica (SiO2) NMs

Silica NMs have diverse applications, such as dye doped fluorescent probes in

nano-bio-imaging [221, 222] and in drug delivery [223]. They are highly hydro-

philic and possess a good potential for surface modifications [222]. The main

mode of action in agricultural formulations is taken over from pharmaceutical ap-

plications, where mesoporous silica NMs are used as controlled release carriers

in drug delivery [224]. Mesoporous silica NMs are solid materials, which consist

of a honeycomb-like porous structure with hundreds of empty channels (meso-

Chapter 2

70

pores) that are able to be loaded with relatively large amounts of active sub-

stances. The unique properties, such as high surface area (>900 m2 g-1), large

pore volume (>0.9 cm3 g-1), tunable pore size with a narrow distribution (2-10

nm), and good chemical and thermal stability of these materials make them po-

tentially suitable for various controlled release applications [223]. Li et al. [225]

synthesized such mesoporous silica NMs as controlled release carriers for the

insecticide Avermectin. With a larger shell thickness the particles could also act

as UV-shields. Increasing the shell thickness in a range of 5–45 nm also led to a

more sustained release of the PPP. Also, these carriers had a high loading ability

for the active substance (approximately 60.0% w/w). However, the efficiency of

this formulation against target organisms was not tested.

In the case of slow release formulations, silica-composite NMs have been pro-

posed. For example, the University of central Florida patented a Silica-Copper

nano-composite formulation [226]. Therein, CuO NMs were loaded onto a silica

nanogel, which is claimed to be formed by “interconnection” of single silica NMs,

to release Cu in a sustained manner and to improve the stickiness to the plant

surface. After two alternating spray applications and drying periods, 41 to 75 % of

the nano-composite were still on the leaf surface, whereas only 1 to 5% remained

of a conventional formulation.

Apart from controlled release, silica NMs have also been proposed as active in-

gredients against insect pests. Since many insects, such as Sitophilus species

that infest agricultural products during storage, have become resistant to a variety

of active ingredients, which also remain as residues on the protected crops,

Debnath et al. proposed the application of surface functionalized silica-NMs as an

insecticide [174] to overcome the problem of resistance development to conven-

tional insecticides. This idea mainly stems from the fact that the insecticidal use

of inert dusts has a long history, from mammals and birds taking “dust baths” up

to civilized nations as well as isolated tribes [227]. The main mode of action of

such NMs against insects is believed to lie in blockage of the trachea or impair-

ment of the digestive tract. Another possible way is the disruption of the lipid-

water barrier that protects insects from desiccation [227]. In the study of Debnath

et al. [174], regardless of the mechanism in action, silica NMs achieved up to


71

69% mortality of Sitophilus adults, whereas the bulk form only reached 23%, indi-

cating the presence of a nano-specific effect.

2.4.2.4 Aluminum (Al) NMs

Aluminum NMs foreseen in agriculture mostly belong to the mineral class of Al-

silicates, as for example kaolin, which occur naturally in soils. Liang and Liu [228]

describe one of the – currently – few nano-fertilizer applications using a polyacryl-

ic acid-co-acrylamide -kaolin nano-composite powder as a slow release carrier

for urea. However, the resulting formulation was not compared to a conventional

urea fertilizer, so no statements on the benefits of use can be made. Another Al-

silicate slow release formulation is described in the Patent US20080194406, filed

by the company Natural Nano Inc [229]. Therein, nanoscale tubular structures

are derived from Al-silicates such as halloysite and serve to a sustained release

of fertilizers as well as PPP. Unfortunately, also in this case, no additional com-

parisons or studies on the efficiency are presented.

Apart from Al-silicates, Stadler et al. [175] proposed the use of “nanostructured

alumina” as insecticides. These Al NMs were based on Al2O3, a versatile ceramic

oxide that has been used in a wide range of applications in electrical, engineering

and biomedical areas [230]. Two species that infest agricultural products during

storage were used as model organisms, similar to the studies of Debnath et al.

[174]: Sitophilus oryzae (L.) and Rhyzopertha dominica (F.). Both species experi-

enced significant mortality after 3 days of continuous exposure to treated wheat.

Nine days after treatment, the LD50 observed ranged from 127 to 235 mg kg−1

[175]. However, mortality due to the Al NMs was only “comparable” to common

insecticidal dusts.

2.4.2.5 Zinc oxide (ZnO)

Zinc oxide is a wide band gap piezoelectric semiconductor with many possible

different morphologies such as rods, wires, sheets and also hollow microspheres

[231]. Zinc oxide NMs have also been found to have superior UV blocking proper-

ties compared to their bulk substitute. This is one of the reasons why ZnO is often

used in the preparation of sunscreen lotions [232]. Its UV-blocking properties are

also one of the main modes of action in agricultural formulations. In patents of

BASF (Patent WO/2009/153231 [154]) and Evonik-Degussa (Patent

Chapter 2

72

WO/2007/014826 [209]), ZnO is proposed as an alternative UV protection agent

to the surface coated TiO2 NMs. Zinc oxide has also been investigated as an ac-

tive ingredient for PPP: Goswami et al. [172] used ZnO NMs as “dust” insecti-

cides, similar to the study of Debnath et al. [174], but found them to be less active

against Sitophilus oryzae than the above described nano-silica (33-65 % mortali-

ty), indicating a material dependency of the effect.

Zinc oxide in its bulk form is also often incorporated into macronutrient fertilizers.

However, the effectiveness of such fertilizers in providing plants with Zn in a Zn

deficient soil is mostly governed by the solubility of the Zn source in the soil. As –

in theory – solubility of particles is depending on particle size, Milani et al. [233]

very recently investigated Zn solubility and dissolution kinetics of ZnO NMs and

ZnO bulk particles coated onto selected granular macronutrient fertilizers. Sur-

prisingly, the kinetics of Zn dissolution was not affected by the size of the ZnO

particles applied for coating, possibly because solubility was controlled by for-

mation of the same compounds irrespective of the size of the original ZnO parti-

cles used for coating.

2.4.2.6 Copper (Cu) NMs

Copper has a long history as a fungicide, especially in vineyards and organic

farming, where it is still used. The Cu ion (redox Cu2+ and Cu1+) is responsible

for the toxicity affecting the activity of several enzymes, thereby preventing ger-

mination of fungal spores. However, this application suffers from the high

amounts of Cu that have to be applied (in the range of 500 to 1500 g ha-1). In the

BASF patent WO/2011/067186 [176] nano-scale amorphous (i.e., non-crystalline)

Cu-salt particles are mixed with a polymer to secondary particle sizes between 1-

200 nm. In one example, such a mixture contained 6.5 g L-1 Cu as Cu-acetate,

sodium acetate and polycarboxylate with particle sizes at 20 nm. One common

non-nano Cu-formulation is Cuprozin® (Spiess Urania Chemicals) that contains

300 g L-1 Cu as Cu-hydroxide. Compared to this formulation, at the same applica-

tion dose of 150 ppm Cu, the nano-formulation could achieve an increase in effi-

ciency of 8% against a phytopathogenic fungus on vines. The use of such a

nano-formulation could thus reduce the amount of Cu introduced into agro-

ecosystems.


73

2.4.2.7 Multiwalled carbon nanotubes (MWCNTs)

Multiwalled carbon nanotubes are NMs composed of multiple layers of extensive

sp2 carbon atoms arranged in fused benzene rings (graphene). Their structure

leads to exceptional electrical, chemical and physical properties, which are in turn

utilized in various applications [234] where unique structural, super conductivity,

and mechanic properties as well as outstanding thermal and chemical stability

are necessary. They have also been proposed as a potential sorbents to remove

organic contaminants due to their relatively high sorption capacity [235].

The first to report positive effects of MWCNT on crops was the group of Khoda-

kovskaya et al. in 2009 [159]. Using an exposure concentration of 10 to 40 µg

mL-1, they observed increased germination rates in tomato seeds. The whole

procedure has very recently been patented in WO/2011/059507 [236]. The main

mode of action is believed to lie in a mechanical penetration of the seed coat,

thus enhancing water uptake and promoting seed germination. This publication

has received a lot of attention in the public (search term “carbon nanotubes toma-

to seeds”, 10.800 hits on 07.05.2012), and its findings were partially inflated into

the grotesque, with open-source article titles such as “CNTs are super-fertilizers”,

“CNT make great tomatoes”, “Want to growth über-Tomatoes really fast? Use

CNTs as fertilizer” etc.

A similar experiment, but addressing growth rate, was conducted recently by

Tripathi et al. [237] using water soluble-CNTs (wsCNTs) as promoters of water

uptake. A concentration of the wsCNT of up to 6.0 µg mL-1 was already sufficient

to enhance the growth rate of common gram (Cicer arietinum) plants. Also, they

provided evidence by optical, fluorescence, SEM and TEM microscopy that these

wsCNT were taken up by plants. Uptake was explained by comparing the outer

diameter of wsCNT (10–30 nm) with that of the xylem (a few microns). Thereby,

they presumed that the wsCNT may get introduced inside the lumen of tracheal

elements. Due to the size difference between these two, wsCNT could be incor-

porated within the xylem according to the concept of the formation of a ‘‘large

capillary’’. These wsCNTs can act then to form several new capillaries which in-

crease the water uptake potential of the plant in addition to the natural flow [237].

They claim that this could be especially useful to enhance water transport and

prevent water loss in agriculture. However, such CNT applications would also

cause dramatically increased fluxes into the soil, compared to current estimations

Chapter 2

74

from unintentional release (Table 2.1). Organic contaminants have a high affinity

to carbonaceous sorbents [238] such as CNT, and several studies postulate that

CNT loaded with them could mediate a “carrier effect” [239] that may in turn lead

to an increase in exposure to environmental micropollutants [240]. Therefore,

such applications should be evaluated with great care.

2.5 Research priorities for a safe use of NMs in PPP and fertiliza-

tion

Intentional and enhanced input of NMs into agricultural ecosystems (Table 2.1) is

an intrinsic prerequisite for “nano-improvement” of PPP and fertilizers, yet many

questions regarding the fate of these materials in the environment still have to be

urgently answered. Despite the fact that soil is the most important environmental

compartment for agricultural production, very few studies investigated the behav-

ior (e.g., mobility and stability) and the effects of NMs in natural soils under envi-

ronmental conditions. For example, size, charge and agglomeration rate of Al

NMs allowed prediction of their mobility in soil [241]. Titanium dioxide NMs easily

passed columns with coarse soils and solutions of low ionic strength, but were

significantly retained in soils with higher clay contents and salinities [242]. Carbon

nanotubes were effectively retained by the soil matrix irrespective of the chemis-

try [243], but may still undergo preferential flow under real conditions and were

shown to be mobile in porous media when associated with natural organic matter

[244].

Only a few efforts have been made to date to assess NM effects on ecosystem

services provided, e.g., by beneficial soil microorganisms. Silver NMs for exam-

ple were able to decrease mycorrhizal colonization of Helianthus annuus [245],

TiO2 and ZnO NMs negatively affected soil microbial communities [246], and sub-

lethal concentrations of quantum dots stimulated genes associated with nitrogen

cycling for example in Azotobacter vinelandii [247]. Understanding the parame-

ters governing NM mobility and effects on organisms in soils is essential for both

risk assessment and design of sustainable agricultural NM applications.

From soils, NMs may be taken up by plants. The current state of research has

been reviewed by Rico et al. [248] and was found to be yet in its beginning. Con-


75

tradicting results were found, and the differences might depend on several fac-

tors, such as species dependent effects, plant age and developing stage, NM

concentration, surface area, and many more. For example, an uptake of <5 nm

TiO2 NMs that were conjugated with alizarin red by Arabidobsis thaliana was re-

ported by Kurepa et al. [249]. The NMs were distributed and accumulated in sub-

cellular regions. In contrast, no uptake of TiO2 NMs in maize plants (Zea mays)

was found by Asli and Neumann [250]. Nanoparticles used in the latter study

were uncoated P25 (Degussa, Germany), and one may assume that a specific

coating might facilitate uptake, for example by interaction with the mucilage sur-

rounding the root. Regarding effects of NMs on agricultural crops, also only few

studies under environmental conditions exist. For example, TiO2 NMs negatively

affected the growth of wheat and soil enzyme activities in a lysimeter study [251].

Suitable analytical methods are lacking to quantify NM concentrations in water,

soil, and air. This hampers risk assessment, where precise data on predicted en-

vironmental concentrations of relevant NMs is essential to define exposure.

Some efforts have been made to solve this gap by modeling environmental con-

centrations [26, 156]. Several analytical techniques have been recently reviewed

by von der Kammer et al. [81]. General analytical difficulties include extraction

and separation/purification from the soil matrix and interfering constituents, as

well as low analyte concentrations. Analysis of metallic NMs is especially difficult,

as many of them, such as TiO2, have naturally occurring counterparts. However,

promising techniques include hyphenated approaches, such as field-flow frac-

tionation (FFF) coupled to ICP-MS, as well as X-Ray based techniques. For car-

bonaceous NMs such as CNTs, further development of chemo-thermal oxidation

[93, 94] and thermogravimetry-mass spectrometry [96] merits additional efforts.

In a classical risk assessment of organic compounds used as PPP, the PBT (per-

sistence, bioaccumulation, toxicity) properties are evaluated. The evaluation of

organic compounds is based on specific endpoints and parameters obtained from

laboratory or field experiments. Persistence is evaluated considering the dissipa-

tion of 50% of the initial concentration, a decision on bioaccumulative properties

of a compound is based on the octanol-water partition coefficient (Kow) and to

evaluate the toxicity criterion, the aquatic toxicity test results determining the in-

trinsic toxicity of the compounds are examined.

Chapter 2

76

It is questionable if these procedures hold true for NMs, since for example Kow is

not a suitable parameter to describe bioaccumulation for NMs. Also in some cas-

es, the determination of the intrinsic toxicity is difficult, especially for compounds

which strongly agglomerate, so that test designs have to be adapted to the prop-

erties of the various NMs. One parameter that could serve this purpose, however,

is the reactivity of the NMs, which is likely to be requested from regulatory bodies

in the future [252]. This parameter could be rapidly determined and emphasize a

need for regulatory action. For example, non-solid NMs such as emulsions or

polymers have been classified as low redox-or catalytically active, whereas solid

metallic NMs such as uncoated TiO2 have been classified as highly active. How-

ever, these classifications still need a refinement, as the determining methods are

not yet standardized.

2.6 Nano-regulation and on-going activities

The fast development of manufactured NMs and their presence on some markets

make it necessary to evaluate their environmental and health impacts. Due to

safety concerns about some NMs, and the problem of inappropriate generaliza-

tion owing to the huge range of nanotechnological applications, it is urgently nec-

essary to address this gap in the regulation of NMs. It should be filled by using

the findings of the ongoing projects in toxicity testing, decision making on material

characterization and testing protocols, and exposure and data management of

the Working Party on Manufactured Nanomaterials of the Organization for Eco-

nomic Cooperation and Development, the European Committee for Standardiza-

tion, and the International Organization for Standardization. Furthermore, an ad-

ditional work programme has been launched as part of the European framework

program FP7 NMP (Nanosciences, nanotechnologies, materials & new produc-

tion technologies) [253], named NANoReg. This intended proposal is aiming to

establish collaboration among authorities of the European governments with re-

gard to the knowledge required for appropriate risk management. The goal of

this project is to provide legislators with a set of tools for risk assessment and

decision making for the short to medium term, by gathering data and performing

pilot risk assessment, including exposure monitoring and control, for a selected


77

number of NMs used in products. A second objective is to bring together the ac-

tivities of national authorities responsible for worker protection, public health and

environment and create the basis for common approaches, mutually acceptable

datasets and risk management. At the current state, there has been an approach

for declaration of nano-pesticides in the US by the Environmental Protection

Agency (EPA). In Europe, the European Chemicals Agency is currently reviewing

the guidance documents to help registrants register nanoforms under REACH. It

aims at sharing experiences with stakeholders on the type of information current-

ly submitted by registrants on NMs. Recently, the European Parliament decided

on a new definition of a NM [8] and included this in a replacement of the Biocidal

Products Directive 98/8/EC [254]. The updated directive states that there is a

“scientific uncertainty” about the safety of NMs. Hence, a nano-form of an active

substance will not be included in the approval procedure, if not explicitly men-

tioned. To fill the gap before the current legislation can be adapted to the needs

of NM regulation, a precautionary matrix has been published in Switzerland help-

ing industry in self-controlling the possible impact of NMs during the production

process, taking into account environmental and health aspects [252].

2.7 Acknowledgements

This work is part of the project “Effects of NANOparticles on beneficial soil MI-

crobes and CROPS (NANOMICROPS)”, within the Swiss National Research

Programme NRP 64 "Opportunities and Risks of Nanomaterials”. We thank the

Swiss National Science Foundation (SNF) for financial support.

Chapter 2

78

2.8 Supporting information

Figure S2.1: Patents of nano-Plant Protection Produ cts and fertilizer applications distribute

as visualized between companies, academia and indiv iduals.

Figure S2.2: Particle sizes of nano-Plant Protectio n Product and fertilizer applications. Par-

ticle sizes that were reported without verification in the respective formulation were indi-

cated as “nominal sizes”.


79

Table S2.1: Reviews published on nanomaterials and agriculture and related fields (in order of appeara nce, non-exhaustive).

Title Focus Differs from current review by… Reference

Nanosilica – from medicine to pest

control

Broad, all kind of applications, reviews nanosili-

ca as a nanobiopesticide

Covering one type of NM only, reviews published

literature only (no patents)

[139]

Nanotechnology applications in pollu-

tion sensing and degradation in agri-

culture: a review

Reviews mainly applications in sensors and in

photocatalysis of pesticides

Focused on remediation (not covering PPP and ferti-

lizer applications), reviews published literature only

(no patents)

[203]

Review of health safety aspects of

nanotechnologies in food production

Overview on the current state of risk assess-

ment (RA) and on scientific issues that need to

be addressed for improved RA of NMs in food

production

Not specifically focused on PPP and fertilizer applica-

tions

[151]

Nanotechnology for parasitic plant

control

Reviews “nanocapsules” as herbicides and

“smart delivery systems” such as viruses

Covering one type of NM only (“Nanocapsules”), only

herbicides, one patent cited

[38]

Agrifood nanotechnology: a tiny revo-

lution in food and agriculture

Very broad overview over potential agricultural

NM applications

Does not cover risks, reviews published literature

only (no patents)

[255]

Nano-particles - A recent approach to

insect pest control

Shortly reviews insecticide applications of NMs Only covers insecticides and very few materials,

does not cover potential risks

[137]

Mycogenic metal nanoparticles: pro-

gress and applications

Large section on NM synthesis in fungi. Also

mentions applications of NMs in medicine

Does very briefly cover PPP and fertilizers, reviews

published literature only (no patents)

[256]

Potential applications of nanotechnol-

ogy in the agro-food sector

Reviews a broad variety of applications, such as

food packaging, pathogen detection, delivery

system as food additives

Not focused on PPP and fertilizers, reviews pub-

lished literature only (no patents)

[39]

Nanoparticulate material delivery to

plants

Focused on transport and effects of NMs in

plants, short excursion on NMs in PPP and ferti-

Focus clearly on reviewing uptake and effect studies,

reviews published literature only (no patents)

[136]

Chapter 2

80

lizer applications

Nanotechnology in agricultural diseas-

es and food safety

Very broad Does not cover risks, reviews published literature

only (no patents)

[37]

Perspectives for nano-biotechnology

enabled protection and nutrition of

plants

Very broad, covers some PPP and fertilizer ap-

plications, but also delivery of genetic material,

nanosensors, remediation

Does not cover patents and legislation as well as

risk/risk assessment

[40]

A brief review of the occurrence, use,

and safety of food-related nanomateri-

als

Focused on food processing, packaging and

storage. Provides a ranking approach for quality

assessment of nano-tox studies

Does not cover PPP/Fertilizer applications, does not

cover patents

[257]

Interaction of Nanoparticles with Edi-

ble Plants and Their Possible Implica-

tions in the Food Chain

Reviews NP-plant interactions and toxicity stud-

ies

Reviews published literature only (no patents), not

focused on PPP/Fertilizers

[248]

Nanotechnology in agriculture Short and broad insight into NMs in agriculture

in general

Does not cover risks, reviews published literature

only (no patents), not focused on PPP/Fertilizer ap-

plications

[36]

Role of nanotechnology in agriculture

with special reference to management

of insect pests

Large section on conventional and biological

insect pest control, NMs for pesticide applica-

tions and as antimicrobial agents for plant path-

ogens

Not covering other applications of PPP or fertilization;

reviews published literature only (no patents)

[258]


81

Table S2.2: Nanomaterials in Plant Protection Produ ct and fertilizer formulations:scientific literatur e.

Refer-

ence

Nanomaterial Reported particle

size [nm]

Applic a-

tion

Function of nanomaterial Country Institution Year

[155] TiO2 30 (nom.) Pesticide Additive (Photocatalyst) China

University 2008

[207] TiO2 30 (nom.) Pesticide Additive (Photocatalyst) China

University 2011

[228] Al-Nanoclay 40-100 Fertilizer Additive (Controlled release) China University 2007

[175] Al-oxide n/a Pesticide Active ingredient USA University 2010

[165] Ag-NP 20-30 (nom.) Pesticide Active ingredient USA University 2009

[158] Ag-NP 1.5 (nom.) Pesticide Active ingredient Korea Company 2008

[166] Ag-NP 4-8 (nom.) Pesticide Active ingredient Korea University 2009

[167] Ag-NP 1-5 (nom.) Pesticide Active ingredient Korea Company 2006

[237] CNT 10-30 (OD), 4-6 (ID) Fertilizer Active ingredient India University 2011

[173] SiO2 n/a Pesticide Active ingredient Germany University 2007

[225] SiO2 140-180 Pesticide Additive (Controlled release) China University 2006

[259] SiO2 80 Pesticide Additive (Controlled release) China University 2006

[172] Al,Zn,Ti,Si,S 40-1106 Pesticide Active ingredient India University 2010

[211] TiO2 5 (nom.) Fertilizer Active ingredient China University 2007

[168] Ag-NP Not provided Pesticide Active ingredient Iran University/Research

center

2010

[260] Polymer/Bifenthrin 60-200 Pesticide Active ingredient USA University 2008

[261] Polymer 200-250 Pesticide Additive (Carrier, Biodelivery) France University/Company 2000

[174] SiO2 15-20 (nom.) Pesticide Active ingredient India University 2010

Chapter 2

82

[178] O/W Emulsion / No-

valuron

200 Pesticide Active ingredient Israel University 2010

[170] Ag-NP n/a Pesticide Active ingredient Korea n/a 2011

[169] Ag-NP n/a Pesticide Active ingredient Korea University 2010

[206] TiO2 n/a Pesticide Active ingredient China University 2006

[187] Polymer 69-127 Pesticide Additive (Controlled release) China University 2009

[262] Cu, Fe, Zn n/a Pesticide Active ingredient Russia n/a 2009

[179] Nano-Emulsion 31 (nom.) Pesticide Active ingredient India University 2011

[263] Nano-Emulsion 20-200 Pesticide Additive (Controlled release, stabi-

lizer)

China University 2009

[159] CNT Not provided Fertilizer Active ingredient USA University 2009

[264] Al-nano clay <100 Fertilizer Additive (Controlled release) Sri Lanka University 2011

[188] Polymer <240 Pesticide Additive(Controlled release) China University 2009

[265] Fe <50 Pesticide Additive (Carrier, Biodelivery) Spain University 2008

[266] Polymer 50-1000/2000 Pesticide Additive (Carrier) USA University 2002

[267] Al-nano clay 30-110 Fertilizer Active ingredient China University 2006

[268] CNT 30 (nom.) Fertilizer Active ingredient India University 2011

[193] Polymer 78 Fertilizer Additive (Controlled release) Brazil University 2010

[269] Al-nano clay n.d. Fertilizer Additive( Controlled release) Malaysia University 2002

[270] Polymer 80 ±30 Pesticide Additive (Controlled release) Bulgaria University 2011


83

Table S2.3: Nanomaterials in Plant Protection Produ ct and fertilizer formulations: patents.

Reference Patent No. Nanomaterial Reported part i-

cle size [nm]

Applic a-

tion

Function of nan o-

material

Country Applicant Year

[163] WO/2003/039249

Polymer 50-2000 Pesticide Additive (Dispersing

agent)

Germany

BASF 2003

[157] DE69126275T2 TiO2 200-300 Pesticide Additive (Dispersing

agent)

France Rhône-Poulenc 1998

[154] WO/2009/153231 TiO2, ZnO,CeO2 1-100 Pesticide Additive (UV-

Protection)

Germany BASF 2009

[271] WO/2003/059070 TiO2 3-200 Pesticide/

Fertilizer

Active ingredient Korea Choi et al. 2003

[272] WO/2008/135093 TiO2, ZnO 10-5000 Pesticide Additive (Carrier) Great

Britain

NM Tech Nano-

materials Ltd.

2008

[226] WO/2010/068275 CuO 1-3 (CuO),7

(SiNP)

Pesticide Active ingredient USA University 2010

[176] WO/2011/067186 Cu 1-200 Pesticide Active ingredient Germany BASF 2011

[273] WO/2010/083319 Cu Not provided Pesticide Active ingredient USA DOW Agroscien-

ce

2010

[236] WO/2011/059507 CNT Not provided Fertilizer Active ingredient USA Khodakovskaya

et al.

2011

[274] WO/2011/031487 CdSe-Quantum dots 2-10 Pesticide Additive (Biodelivery) USA University 2011

[180] WO/2010/051607 Nano-Emulsion <400 Pesticide Additive Brazil Oxiteno SA. 2010

[229] US20080194406 Al-nano clay Not provided Pesticide Additive (Controlled USA NaturalNano, Inc. 2008

Chapter 2

84

release)

[182] WO/2008/032328 Nano-Emulsion <300 Pesticide Additive (Dispersing

agent)

Israel University 2008

[171] US20090075818 Ag-NP 4.5 Pesticide Active ingredient Iran Raman Nia, J. 2009

[181] US20100041629 Nano-Emulsion 10-300 Pesticide Additive (Dispersing

agent)

Germany Giessler-Blank et

al.

2010

[194] WO/2010/035118 Polymer <100 Pesticide Additive (Controlled

release)

USA Vive Nano, Inc. 2010

[275] CN1433697 Polymer/Pyrethrines n/a Pesticide Active ingredient China University 2003

[276] US2011000411 Si-clay 1µm-2mm (with

~50 nm Pores)

Fertilizer Additive (Controlled

release)

USA Polymate Ltd. 2011

[277] DE102009030121 Ag, Au, Cu, Zn 1-100 Fertilizer Active ingredient Germany Rent a scientist

GmbH

2010

[278] WO/2006/049379 SiO2-Ag 0.5-30 Pesticide Active ingredient Korea Bio Dreams, Co.

Ltd.

2006

[209] WO/2007/014826 TiO2, ZnO <200

/aggregates

UV-

Protection

of plants

Active ingredient Germany Evonik-Degussa 2007

[279] CN1491551 Polymer/Ivermectin <100 Pesticide Additive (Dispersing

agent)


[280] CN1491558 Polymer/Acetamiprid <100 Pesticide Additive (Dispersing

agent)


[281] WO/2002/082900 Polymer <1000 Pesticide Additive (Controlled

release)

USA Rhodia, Inc. 2002

[282] WO/2008/056234 SiO2 <2000 Fertilizer Additive (Controlled

release)

India Bijam Bioscience,

Ltd.

2008


85

[283] US8017061 Polymer 1000-2000 Pesticide Additive (Controlled

release)

Germany University 2011

[189] US7070795 Polymer 100nm-200µm Pesticide Additive (Controlled

release)

USA Monsanto 2006

[284] US7994227 Polymer <500 Pesticide Additive (Controlled

release)

Germany BASF 2011

[285] WO/2005/115143 Polymer 10-4000 Pestici-

de/Fertilizer

Additive (Controlled

release)

Germany University 2005

[286] WO/2011/053605 Polymer n/a Pesticide Additive (Controlled

release)

USA Dendritic Nano-

technologiesInc.

2011

[192] US7494526 Cochleates (Lipid

vesicles)

<1000 Fertilizer Additive (Controlled

release, Biodelivery)

USA Yavitz, E. 2005

[184] WO2011138701 Nano-Emulsion 50-800 Pesticide Additive (Controlled

release)

Germany BASF 2011

[287] WO2011010910 Nano-Emulsion <100 Pesticide Additive (Biodelivery) Malaysia University 2011

Chapter 2

86

Table S2.4: Nanomaterials in Plant Protection Produ ct and fertilizer formulations: products

Product (Potential) Nanomaterial Reported particle size

[nm]

Application Manufacturer

Primo MAXX (Nano-)Emulsion n/a Pesticide Syngenta

Banner MAXX (Nano-)Emulsion n/a Pesticide Syngenta

Karate ZEON Capsules with λ-cyhalotrin >100 Pesticide Syngenta

Demand CS Capsules with pyrethroid (λ-cyhalotrin) >100 (>1000 [288]) Pesticide Syngenta

ECOFLEX Aliphatic copolyester, "nanofibre" as a

pheromone dispenser

n/a Pesticide BASF

Aerosil 200 SiO2 12 Various purposes Syngenta

Trico TiO2 23.2%, Sheepgrease n/a Pesticide Omya AG

FEROX Zero-valent iron nanopowder 10-100 Example of a Nanomaterial

for soil remediation

ARS Technologies, Inc.

SoilSet SiO2 n/a Example of a Nanomaterial

for soil management

Sequoia Pacific Research Company

87

Potential of Hyperspectral Chapter 3

Imaging Microscopy for semi-

quantitative analysis of nanoparticle up-

take by protozoa

Monika Mortimer§, Alexander Gogos§, Nora Bartolome, Anne Kahru, Thomas D.

Bucheli and Vera I. Slaveykova

Reprinted with permission from: Environmental Science and Technology 2014,

48(15): p. 8760-8767. Copyright 2014 American Chemical Society.

§Authors contributed equally to this work

Chapter 3

88

Abstract

Hyperspectral imaging with enhanced darkfield microscopy (HSI-M) possesses

unique advantages in its simplicity and noninvasiveness. In consideration of the

urgent need for profound knowledge on the behavior and effects of engineered

nanoparticles (NPs), here we determined the capability of HSI-M for examining

cellular uptake of different metal-based NPs, including nanosized metals (silver

and gold, both citrate stabilized), metal oxides (copper oxide and titanium diox-

ide), and CdSe/ZnS core/shell quantum dots at subtoxic concentrations. Specifi-

cally, we demonstrated that HSI-M can be used to detect and semi-quantify these

NPs in the ciliated protozoan Tetrahymena thermophila as a model aquatic or-

ganism. Detection and semi-quantification was achieved based on spectral librar-

ies for the NPs suspended in extracellular substances secreted by this single-

celled organism, accounting for matrix effects. HSI-M was able to differentiate

different NP types, provided that spectral profiles were significantly different from

each other. This difference in turn depended on NP type, size, agglomeration

status and position relative to the focal plane. As an exception from the NPs ana-

lyzed in this study, titanium dioxide NPs showed spectral similarities compared to

cell material of unexposed control cells, leading to false positives. High biological

variability resulted in highly variable uptake of NPs in cells of the same sample as

well as between different exposures. We therefore encourage the development of

techniques able to reduce the –currently- long analysis times that still hamper the

acquisition of statistically strong datasets. Overall this study demonstrates the

potential and challenges of HSI-M in monitoring cellular uptake of synthetic NPs.

Potential of Hyperspectral Imaging Microscopy for semi-quantitative analysis of nanoparticle uptake by protozoa

89

3.1 Introduction

Due to increasing use of engineered nanoparticles (NPs) in consumer products,

one of the major pathways of nanomaterial release to the environment is consid-

ered to be through wastewater systems [289]. NPs entering the environment via

urban effluents is a potential threat to ecosystems. Thus, for complete environ-

mental risk assessment it is important to be able to detect and quantify NPs in

biological samples, including aquatic organisms.

To date the most frequently employed methods for imaging of non-fluorescent

NPs are electron microscopy-based techniques coupled to X-ray spectroscopy.

While electron microscopy has significantly higher resolution than light microsco-

py, its disadvantages include expensive instrumentation and sample preparation

which in addition to being laborious could also introduce artifacts in the speci-

mens [81]. In recent years, alternative methods for imaging of non-fluorescent

NPs have been developed and increasingly used, including methods based on

detection of light scattering by the particles [290] as well as hyperspectral imag-

ing with enhanced darkfield microscopy (HSI-M) [83]. HSI-M with its high intensity

darkfield offers approximately 150 times higher illumination of particles than con-

ventional optical methods, allowing NPs to appear bright and to record full visible

and near-infrared (VNIR) spectra for each image pixel. Spectral profiles of NPs

can be used with software-based spectral classification methods to detect the

NPs in hyperspectral images (HSIs) of biological samples. Depending on the

spectral profiles of NPs, HSI-M offers the possibility of localizing different types of

NPs in heterogeneous samples. Additionally, sample preparation is fast and

easy, reducing the number of potential artifacts that are easily introduced in other

microscopy techniques.

Recently, Badireddy et al. [103] demonstrated the use of HSI-M for detection,

characterization, and analysis of different types of engineered NPs in pure and

complex water samples such as simulated-wetland ecosystem water and

wastewater. In the current study HSI-M was used for the detection and subcellu-

lar localization of silver (Ag), gold (Au), titanium dioxide (TiO2), copper oxide

(CuO) NPs and quantum dots (QDs) in a unicellular freshwater organism - ciliat-

ed protozoan Tetrahymena thermophila. Ciliated protozoa, being widely distribut-

ed and ecologically significant, can be good indicators of harmful effects of engi-

Chapter 3

90

neered NPs in ecosystems. Due to the grazing ability, protozoa can accumulate

considerable amounts of NPs even at low environmental NP concentrations

which could consequently become bioavailable for the organisms at higher

trophic levels. Moreover, protozoan grazing on engineered NP could affect the

planktonic systems through decreased trophic transfer efficiencies and reduced

regeneration of nutrients: studies have shown that exposure to QDs [146] and

TiO2 [291] impaired the bacterial digestion by T. thermophila. However, in general

T. thermophila has been found to be more tolerant to toxicants, including engi-

neered NPs, compared to other fresh water organisms. For instance, the toxic

concentrations (EC50, effective concentration leading to 50% loss in cell viability)

of Ag NPs in aquatic organisms range from 0.01 mg L-1 for crustaceans to

10 mg L-1 for fish but up to 100 mg L-1 in protozoa [292, 293]. Similar trends in

sensitivities of aquatic species have been reported for CuO NPs [292, 294]. Rela-

tively high tolerance for toxicants allows potential application of the protozoans as

accumulating and removing NPs from the environment and also as suitable mod-

els for studying the uptake and clearance of NPs in the cells.

The aim of this study was to demonstrate the potential of HSI-M for semi-

quantitative analysis of NP uptake in the cells. We used protozoa as model cells

for exploring the advantages and limitations of HSI-M in cellular uptake studies of

engineered NPs. We asked whether coating the reference NPs with protozoan

extracellular substances (ES) would change the detection rate of these NPs in

the sample cells, how NP scattering and size/agglomeration characteristics influ-

ence their detection rate and what is the capability of HSI-M in distinguishing NPs

of different materials internalized in the cell. In addition, we propose potential ap-

plications of HSI-M in elucidation of the mechanisms of biological effects of NPs

when used in combination with biological assays.


91

3.2 Experimental

3.2.1 Nanoparticles

Ag and Au NPs – each in two sizes (10 and 20 nm) – were purchased from Sig-

ma-Aldrich (Buchs, Switzerland): Ag NPs were purchased as 0.02 mg mL-1 sus-

pensions in aqueous buffer containing sodium citrate as a stabilizer and Au NPs

in reactant free 0.1 mM phosphate buffered saline (PBS) buffer. CdSe/ZnS

core/shell QDs with terminal carboxyl groups (Qdot® 655 ITK™ Carboxyl Quan-

tum Dots, 8 µM solution in 50 mM borate buffer, pH 9.0) were purchased from

Life Technologies Europe B.V., Carlsbad, USA. CuO (Nanostructured and amor-

phous materials Inc., Huston, TX, USA) and TiO2 (Degussa P25, Evonik Indus-

tries, Essen, Germany) NPs were purchased in powder form and stock suspen-

sions were prepared in MilliQ water at a concentration of 10 g L-1, sonicated in

ultrasonication bath for 30 min and stored at 4°C in the dark. Prior to the experi-

ments the stock suspensions of CuO and TiO2 NPs were diluted in 10 mM 4-(2-

hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES, Sigma-Aldrich, Buchs,

Switzerland) at a pH of 7.0. The average hydrodynamic diameter and zeta-

potential of NPs in HEPES buffer were measured using a Zetasizer Nano ZS

(Malvern Instruments, Renens, Switzerland). The characteristics of the NPs used

in the current study are shown in Table 3.1.

Chapter 3

92

Table 3.1: Characteristics of nanoparticles.

Nano-

particle

Size,a

nm

Mass conc.,

mg L -1

Particle

num-

ber/L a

Hydrodynamic

diameter, b nm

ζ-potential, b

mV

Hamaker

constant,

A121

(X10-20J)c

Au 10 25 ~3×1015 15.8±0.5 -33.6±0.8 27

Au 20 25 ~3.5×1014 27.5±0.3 -39.2±0.3 27

Ag 10 10 ~2×1015 19.6±0.8 -39.6±0.1 28.2

Ag 20 10 ~2.3×1014 25.7±0.4 -46.1±4.7 28.2

TiO2 20 10 NA 1840±170 -12.2±0.1 6

CuO 30-50 10 NA 778±101 -30.0±0.9 0.54g

CuO 30-50 50 NA 704±3.5 -15.6±2.2 0.54g

QDs 10 2.9d 20e 12.8±0.5f -21.5±6.0 7.78h

aData provided by the manufacturer. bMeasured with Zetasizer, stock suspensions of NPs dilut-

ed with 10 mM HEPES buffer, pH 7.0. Values are means of 3 measurement ± standard devia-

tion. Measurements were performed using samples freshly prepared from stock dispersions. cFrom Petosa et al., 2010 [295] if not stated otherwise. dmg of total Cd/L. enM, data provided by

the manufacturer. fCorresponds to the AFlFFF peak maximum. gFrom Jeong and Kim, 2014

[296]. hFrom Morrow et al., 2006 [297]. NA – data not available

3.2.2 Protozoan culture

Protozoan culture (Tetrahymena thermophila strain BIII) was cultivated in a pro-

teose peptone based medium, supplemented with 250 µg mL-1 of each penicil-

lin G and streptomycin, and 1.25 µg mL-1 amphotericin B (all from Sigma-Aldrich)

following the procedure described previously [294]. At exponential growth phase

(5×105 cells mL-1) protozoa were harvested by centrifugation (850g, 5 min, 4°C)

and washed twice with 10 mM HEPES buffer, pH 7.0. Cell numbers were deter-

mined using a Coulter Multisizer III particle counter (Beckman-Coulter, Switzer-

land).

3.2.3 Exposure of protozoa to NPs and preparation of the cells for microscopy

All the exposures of the cells to NPs were done in 10 mM HEPES buffer, pH 7.0

at 25°C in the dark. Protozoan culture at twice the final cell concentration

(5×105 cells mL-1) was mixed in 1:1 ratio with NP suspension or HEPES buffer

for unexposed controls. The concentrations of NPs used in the uptake experi-


93

ments are shown in Table 3.1. In addition, Au NPs were also used at 10 mg L-1

and TiO2 NPs at 100 mg L-1. These concentrations of NPs were chosen because

in this concentration range protozoa accumulated sufficient amounts of NPs for

detection by optical microscopy while still remaining at sub-lethal level to the test

organisms. Au, CuO and TiO2 NPs were used at two different concentrations in

exposure experiments to explore concentration dependent uptake of different

types of NPs. Co-exposure to Ag NPs and QDs was done using concentrations of

5 mg L-1 and 5 nM (0.725 mg of total Cd L-1), respectively, to avoid synergic toxic

effect of NPs. T. thermophila was incubated with or without NPs either for 2 or

24 h, then harvested by centrifugation at 850g for 5 min at 4° C and resuspended

in 10 mM HEPES (pH=7.0). The cells were fixed in 4% paraformaldehyde for

15 min at room temperature, the fixative was removed by centrifugation and the

cells were resuspended in HEPES buffer. 5 µL of cell suspension was pipetted on

a glass slide, mixed with 5 µL of Mowiol® 4-88 (Polysciences Europe GmbH, Ep-

pelheim, Germany), covered with a glass slip, allowed to harden at room temper-

ature for 24 h and kept at 4° C until imaging.

3.2.4 Preparation of reference samples

Two types of reference samples were used in the current study: (i) 5 µL of each

of the studied NP suspension was pipetted on a glass slide, mixed with 5 µL of

Mowiol® 4-88, covered with a glass slip, allowed to harden at room temperature

for 24 h and kept at 4° C until imaging; (ii) T. thermophila was incubated in

10 mM HEPES, pH 7.0, for 2 h to allow synthesis of extracellular substances

(ES) [298], the cells were separated from ES fraction by centrifugation at 850g for

5 min, 4° C, 100 µL of ES was mixed with 100 µL of NP suspension and incubat-

ed at 25°C in the dark for 22 h. 5 µL of ES incubated NP suspension was used

for microscope sample preparation as described above. Embedding the refer-

ence NPs in Mowiol was necessary to ensure identical imaging conditions with

the samples and because of an inherent limitation of HSI-M in imaging non-fixed

liquid samples, where Brownian and diffusive movement of the particles render

HSI image acquisition impossible. However, most of the reference particles were

discernible as individuals when fixed in Mowiol (Figure S3.1).

Chapter 3

94

3.2.5 Hyperspectral imaging

Samples were analyzed using an enhanced darkfield transmission optical micro-

scope (Olympus BX43) equipped with dual mode fluorescence (DMF) and a HSI

spectrophotometer (CytoViva Inc, Auburn, USA). The DMF module allowed ob-

serving the sample while switching between fluorescence and non-fluorescence

modes. The HSI spectrophotometer was used for recording spectra with a low

signal to noise ratio in visible and near-infrared (VNIR) wavelengths (400 –

1000 nm) at a high spectral resolution of 2.5 nm. These scans resulted in 3d-

structured data cubes, which consisted of a 2d-image with a full spectrum in each

pixel with a size of 90 × 90 nm. These data cubes are referred to as hyperspec-

tral images (HSIs) throughout this paper. For the semi-quantification analysis, 10

single cell images were collected randomly for each type of NP exposure (sample

HSIs).

3.2.6 Image processing and analysis

HSI analysis was done using Environment for Visualizing Images (ENVI) 4.8

software (Exelis Visual Information Solutions, Inc., USA) that contained special

image analysis features added by CytoViva, including normalization for lamp

spectrum, adjacent band averaging and particle filter analysis. A detailed descrip-

tion of all executed steps for HSI acquisition can be found in the supporting in-

formation (SI). All HSI analysis results in this study were obtained after executing

the steps described therein. Briefly, and in general (see Figure 3.1 for schematic

process illustration), before HSI analysis of the specimen, a spectral library (SL)

is created from the NPs alone, consisting of representative spectral profiles.

Then, an HSI is recorded from the specimen (e.g. the protozoan cell) and a spec-

tral classification algorithm (spectral angle mapper, SAM) is applied on the spec-

imen HSI using the SL created beforehand to map spectral similarities in the HSI.

The result of this mapping process is either an image in which the matching pix-

els are highlighted, or a quantitative measure with a percentage of mapped pixels

per image. HSI results were additionally normalized to cell area by manual

measurement in ImageJ (NIH, USA).


95

Figure 3.1: A scheme of hyperspectral image (HSI) a cquisition and analysis of NP-exposed

protozoan cells. NPs - nanoparticles; SL - spectral library; SAM - spectral angle mapper.

3.2.7 Statistical analysis

Statistical analysis was performed using analysis of variance (ANOVA). Calcula-

tions were made using the software R (R version 3.01, the R foundation for statis-

tical computing) integrated in RStudio (version 0.97.551). For each treatment, 10

cells were analyzed (n=10), except for the Au NP treatments that were repeated

twice (n=20). For pairwise comparison of different parameters (e.g. 2h vs. 24h)

an F-test was used.

Chapter 3

96

3.3 Results and discussion

3.3.1 Influence of protozoan extracellular substances (ES) on visible and near

infrared (VNIR) spectral profiles of nanoparticles and the characteristics

thereof

In the current study Au and Ag NPs of two primary sizes (10 and 20 nm), TiO2

and CuO NPs as well as QDs were employed (Table 3.1). Both Au and Ag NP

suspensions were stable in HEPES buffer as indicated by their high negative ζ-

potentials while their hydrodynamic diameters were approximately 5 to 10 nm

larger than their primary sizes provided by the manufacturer. Conversely,

HEPES-suspended TiO2 and CuO NPs showed high levels of aggregation and

the respective hydrodynamic diameters were in the range of micrometers (Table

3.1). Different agglomeration patterns of metallic and oxide NPs can be explained

by their ζ-potential and Hamaker constant values. Protozoa are known to excrete

extracellular substances [293, 299, 300]. It has been shown that ES excreted by

T. thermophila are composed mainly of carbohydrates and nucleic acids [298] but

also of more than 30 different proteins, mostly proteases [293, 300]. In our exper-

iments we tested the influence of ES on the VNIR spectral profiles of the NPs. To

this end, we extracted SLs from NPs in HEPES as well as NPs incubated with

protozoan ES (for detailed steps, see SI). Based on observations with the naked

eye, no apparent agglomeration or precipitation of NPs was detected during in-

cubation with ES, i. e. visually the color and appearance of the dispersions did

not differ from that of the NP dispersions without ES, whereas HSI-M images and

spectra were partially different.

Reference spectra were collected from the NP suspensions which were prepared

by directly suspending the NPs in aqueous media (Figure 3.2, black lines) or after

incubating NPs with protozoan ES (Figure 3.2, red lines). The spectra of Ag and

Au NPs both showed primary size-dependence with peak maxima of 10 nm-sized

particles at lower wavelengths than 20 nm-sized NPs (Figure 3.2A, B, C, D). This

is in accordance with available literature for Ag NP [301], where the spectral peak

for smaller particulate Ag (2-11 nm) was around 500-550 nm.

When NPs were incubated with protozoan ES the reflectance spectra of all NPs

except for TiO2 showed a shift compared to the spectra of NPs in aqueous media

(Figure 3.2). The spectra of Ag and CuO NPs red-shifted after incubation with


97

ES. Red-shifting of the Ag resonance peak has been associated with particle ag-

gregation in case of Ag NPs [302] and changes in the surface properties of Ag

NPs [303], indicating possible agglomeration of NPs facilitated by ES or NPs be-

coming coated with the biomolecules excreted by protozoa. Incubation with pro-

tozoan ES has been shown to increase the hydrodynamic diameter of QDs [304].

Interestingly, an opposite effect occurred in the case of Au NPs, with the spectra

blue-shifted for both 10 and 20 nm-sized particles. Larger blue-shifts occurred in

the case of 20 nm than the 10 nm-sized Au NPs. Blue-shifting could suggest that

ES promoted the suspendability of individual Au NPs to decrease their average

size distribution [305]. Such blue-shift in VNIR scattering spectra has been re-

ported upon Ag NP binding with actin which facilitated the breakdown of Ag NP

aggregates [306]. Slight blue shift in localized surface plasmon resonance (SPR)

peak of Ag NPs has also been associated with SPR masking by the NP surface

bound organic ligands [303]. Shifts in opposite directions were induced in VNIR

spectra by ES, i.e. blue-shift in case of Au NPs and red-shift in case of Ag and

CuO NPs, indicating the complex roles of ES on NP aggregation and stabiliza-

tion, which depend on both the chemical compositions of the ES as well as the

physicochemical properties of the NPs [307, 308]. In the case of Ag NP, ES incu-

bation also induced changes in the peak morphology, with a reduction of the

peak shoulder in the 500 nm region, which was more pronounced in the 20 nm

sized particles possibly due to NP-ES complex formation [309].

Chapter 3

98

Figure 3.2: Spectral signatures of reference NPs (s ee Table 3.1). Black lines represent the

averaged spectra of the reference NPs in aqueous su spensions and red lines are the aver-

aged spectra of NPs incubated in protozoan ES.

3.3.2 Spectral angle mapping of nanoparticles in protozoa

In the current study freshwater ciliate T. thermophila was used as a model organ-

ism for examining the uptake of NPs by HSI-M. As a phagocytizing organism T.

thermophila is known to ingest NPs into its food vacuoles [294, 304, 310]. In our

experiments, incubation with Au, Ag, CuO and TiO2 NPs up to 24 h had no effect

on protozoan viability compared to control cells with the exception of TiO2 NPs

which caused some mortality (15%) after 24 h exposure to 10 mg TiO2 L-1 but not

after 2 h exposure (Figure S3.2). The relatively low level of 15% lethality after 24h

exposure of protozoa to 10 mg L-1 TiO2 could be caused by the retention of TiO2

in the cells and the resulting oxidative stress related toxicity as has been reported

in the literature [311]. During the 2h exposure protozoan food vacuoles became

filled with aggregates of NPs as seen in brightfield (BF; Figure S3.3) and in HSI-

M (Figure 3.3) images. Internalization of NPs was confirmed by observing differ-

ent focal planes (3-5); NP aggregates were sharp and in focus together with the

surrounding cell wall only in the middle 1-2 planes (Figure S3.4) and thus in the

center of the organism. This approach has also been used in other recent studies

[312].


99

Internalized NPs were detected by SAM analysis on the respective HSIs. Initially,

SAM analysis was performed using comparatively SLs from NPs in aqueous sus-

pensions and SLs obtained from NPs incubated in protozoan ES. The results

showed that, specifically in the case of protozoa exposed to 20 nm Ag NPs, using

SLs from NPs incubated in ES as references SAM analysis yielded higher num-

ber of matched pixels in the cells (Figure S3.5). It can thus be assumed that the

spectral libraries collected from the 20 nm reference Ag NP incubated with ES

were more representative of the particles internalized in the cells. Although the

influence of spectral shift upon NP incubation with protozoan ES had a significant

influence on SAM analysis of the NP content in the cells only in case of 20 nm Ag

NPs (Table S3.1), similar tendencies were noted with other NPs. Therefore, in

the following, only HSI results obtained from SLs of NPs with ES are shown and

discussed.

Figure 3.3 shows localized spectral profiles of the NPs inside the protozoan cells

detected by SAM analysis. In general, NPs and their agglomerates that showed

too high/too low intensities were not detected by HSI-M (10 nm Au NPs for ex-

ample, Figure 3.3C). This phenomenon was size dependent and thus prominent

in samples where particle size was multi-modal. In such, larger agglomerates

scattered at very high intensities, whereas smaller ones showed very low scatter-

ing. In addition, highly defocussed particles showed lower intensities and in-

creased noise. At intensities that exceed the dynamic range of the detector, peak

deformation occurs, leading to non-recognition of these pixels during SAM analy-

sis. At very low intensities, NP pixels might have not been recognized by SAM

due to decreased signal to noise ratio. Increased detection rate of NPs could be

obtained by generation of “high definition range”-analogue HSIs at multiple in-

creasing/decreasing exposures and subsequent allocation of the different match-

ing results. However such technical development was beyond the scope of the

current study.

In the studied system, all spectra except for TiO2 were highly specific for the stud-

ied material, as no false positives were detected in unexposed control cells. In

case of TiO2 SAM analysis of control cells led to false positives, detecting NP

spectral profiles in the cells (Figure S3.6). Also in some cases (see darkfield im-

age of 10 mg L-1 TiO2, Figure 3.3G), cells obviously contained TiO2 NP, as was

evident also from the BF images (Figure S3.3) but were not recognized by SAM

Chapter 3

100

as such. Thus, cell material may not provide enough contrast for specific detec-

tion of TiO2 using HSI-M.

Figure 3.3: Representative hyperspectral images of Tetrahymena thermophila exposed to

NPs for 2 h. (A, B, C, D) Protozoa incubated with 1 0 nm Ag NPs, 20 nm Ag NPs, 10 nm Au

NPs and 20 nm Au NPs, respectively, at 10 mg L -1. (E, F, G, H) Protozoa exposed to 10 mg

L-1 and 50 mg L -1 CuO NPs and 10 mg L -1 and 100 mg L -1 TiO2 NPs, respectively. Red dots

on the dark field images represent spectral angle m apper (SAM)-localized spectral profiles

of the respective NPs. All images were acquired usi ng 100× objective, scale bar = 10 µm.

3.3.3 Nanoparticle differentiation capabilities of HSI-M

To further explore the detection and NP differentiation capabilities of HSI-M, we

conducted an experiment where protozoa were exposed to a mixture of two types


101

of NPs. For this, we chose fluorescent CdSe/ZnS core/shell QDs to be able to

verify HSI-M results with fluorescence imaging and 10 nm-sized Ag NPs. QDs

and Ag NPs were chosen for co-exposure because their VNIR spectral peaks

were at a considerably different wavelength (547 and 660 nm, respectively; Fig-

ures 3.2A and S3.7) which represented an ideal case for simultaneous detection

of the two signals. The scattering spectrum of QDs corresponded well to the

spectrum of the same type of QDs reported in the literature with the peak maxi-

mum at 660 nm [103]. To avoid a synergic toxic effect of the NP mixture, the cells

were exposed to lower concentrations of each NP than in the single NP exposure

experiments (5 mg L-1 of Ag NPs and 5 nM QDs). Figure 3.4 shows a representa-

tive image of a cell, where QDs were detected inside the cell by fluorescence im-

aging (Figure 3.4A). HSI analysis was performed on the same cell, revealing QD

signatures at the same areas (Figure 3.4B). Ag NP spectral signatures were de-

tected by SAM at different locations in the same cell (Figure 3.4C). However,

scattering intensities of the two particles were very different due to higher ag-

glomeration of the QDs compared to the AgNP. To account for these differences,

the spectral angle used for SAM analysis was modified separately for each NP

type. In case of QDs a wider spectral angle (0.6) compared to Ag NP (0.2) was

used for detection.

SAM analysis of the mixed NP exposure samples showed that optimization of the

spectral angle allowed increased detection of NPs in biological samples. In addi-

tion, HSI-M enabled differentiation of different types of NPs in cells in cases

where the spectral profiles of the NPs were clearly distinguishable from each oth-

er. To illustrate this we performed SAM analysis on HSIs of Ag-incubated cells

using spectral libraries (SLs) of Au NP reference particles and Au-incubated cells

using SLs of Ag NP reference particles (Figure S3.8). When NP reference spec-

tra were different enough (in case of Ag 10 nm and Au 20 nm, Figure 3.2), no

false positives were observed, whereas very similar NP spectra (e.g., 20nm-sized

Ag and Au NPs, Figure 2) resulted in the detection of false positive matched pix-

els in the sample cells (Figure S3.8). Thus, depending on the NP type, size, ho-

mo-/heteroagglomeration status and position relative to the focal plane, HSI-M

appears to have the capacity to provide specific chemical data on selected NPs

in biological systems.

Chapter 3

102

Figure 3.4: Single-cell analysis of a mixed exposur e experiment of Tetrahymena ther-

mophila to fluorescent CdSe/ZnS core/shell QDs (5 nM) and Ag NPs (5 mg L -1). (A) Fluores-

cence image of NP exposed cell and (B) hyperspectra l image (HSI) of the same cell with

spectral angle mapper (SAM) matching result for QD spectral profiles at a spectral angle of

0.6 radians. (C) HSI of the cell with SAM matching result for Ag NP spectral profiles at a

spectral angle of 0.2 radians. Red dots on dark fie ld images (B and C) represent localized

spectral profiles of NPs and are highlighted with a n arrow. Red fluorescence of QDs in (A)

is visible in the same regions recognized by HSI in (B).

3.3.4 Semiquantitative characterization of nanoparticle uptake and clearance by

T. thermophila as measured by HSI-M

Protozoa generally contained highly variable amounts of NPs/NP agglomerates

even in the same sample with some cells having food vacuoles filled with NPs

and the other cells containing insignificant amounts of NPs (Figure S3.9). How-

ever, with imaging techniques such as HSI-M and BF imaging it was possible to

assess the cellular uptake rate of NPs semi-quantitatively. The more NPs were

taken up, the more NPs were detected by the two methods independently (see SI

for detailed description of BF microscopy analysis and Figure S3.10 for correla-

tion between the two methods).

In BF, NP agglomerates appeared as dark areas in the cells, and thus could be

believed to correlate with the amount of NP taken up or cleared during exposure.

However, in BF microscopy resolution is restricted to ~0.2 µm which allows imag-

ing and analysis of only the agglomerates of NPs. HSI-M in turn holds the poten-

tial of additionally capturing spatially resolved single NPs and very small agglom-

erates by its enhanced darkfield as well as providing information on the composi-

tion of the NPs by spectral analysis.


103

Overall, the measured percentage of matched pixels per cell area (Figure 3.5)

showed a high variability with relative standard deviation (RSD) varying between

29 and 227%. One of the likely reasons for this is biological variability in terms of

NP uptake by protozoa. In most cases, the 2-h samples showed a lower RSD

compared to the 24-h ones. This is explainable by the fact that after 24-h expo-

sure the protozoa contained less NPs than after 2-h exposure, including some

cells which were completely devoid of NPs after 24 h leading to the matching val-

ue of zero. Also, the cellular content of 20 nm-sized Ag NPs showed lower RSDs

(represented by whiskers in Figure 3.5A) compared to that of 10nm Ag NPs. This

could be related to the higher dissolution rate of the latter due to smaller size and

larger surface areas for interacting with cellular molecular contents as 10 nm Au

NPs at the same concentration were present in the cells in much higher amounts

(Figure 3.5B).

In general, when analyzed over the whole dataset ANOVA revealed that NP type,

concentration, size and exposure time had a significant influence on HSI-M re-

sults (Table S3.1). In all cases, when compared overall, NP contents in the cells

after 24 h were significantly lower (P<0.001) compared to 2-h exposures. Proto-

zoa are known to graze unselectively on a range of nano-and micro-sized parti-

cles in the surrounding medium and internalize particles up to a few micrometers

[313]. The clearance of the cells after 24-h incubation with NPs could be caused

by the effective grazing ability and fast turnover of food vacuoles in protozoans,

which leads to formation of stable NP-biomolecule agglomerates excreted from

food vacuoles and too large to be re-ingested, a phenomenon clearly visible in

BF images (Figure S3.9) and also reported in our study with T. thermophila and

QDs [304]. Uptake rates of Au NPs were independent of primary particle size with

no statistical difference in uptake of 10 and 20 nm Au NP (P>0.05). Likewise,

there was no significant difference (P>0.05) in 2-h uptake of 10 and 20 nm Ag

NPs but after 24 h the 20 nm Ag NP exposed cells contained significantly higher

amounts of NPs than the cells exposed to 10 nm Ag NPs (P<0.05), likely due to

their differential dissolution [314]. A general, however, not statistically significant,

trend of 10 nm-sized Ag and Au particles being taken up during 2 h exposure at a

greater extent than 20 nm particles was noted (Figure 3.5A, B, C). This is likely a

result of the faster diffusion and approximately ten-fold higher particle number of

the 10 nm than the 20 nm Ag and Au NPs at equal mass concentrations (Ta-

Chapter 3

104

ble 3.1). Both CuO and TiO2 NPs showed a tendency of concentration dependent

uptake (10 vs. 50 mg L-1 and 10 vs. 100 mg L-1, respectively, Figure 3.5D, E).

However, after 24 h the intracellular content of CuO NPs was not significantly

different for 10 and 50 mg L-1 exposure. NP clearance at 24 h was the lowest in

case of exposure to Au NP at 25 mg L-1 and TiO2 at 100 mg L-1 (Figure 3.5C, E).

In general, NPs that are known not to dissolve (Au, TiO2) showed a highly signifi-

cant (P<0.001) elevated uptake at the same exposure concentrations compared

to NPs with higher dissolution rates (Ag and CuO), which could indicate a possi-

ble inhibitory effect of dissolved copper and silver ions to protozoan grazing activ-

ity.


105

Figure 3.5: Boxplots showing semiquantification res ults of internalized NPs by Tetrahy-

mena thermophila after incubation for 2 and 24 h, presented as % ma tched pixels per cell

area. Semiquantification was done using reference s pectra obtained from extracellular

substances (ES) incubated NPs. Horizontal lines den ote mean values and black dots are

outliers. Whiskers signify standard deviations (n=1 0).

Chapter 3

106

3.4 Application potential of HSI-M and outlook

In this study we have demonstrated the application of HSI-M for intracellular de-

tection and identification of different metal-based NPs, including differentiation of

two types of NPs internalized in the cell simultaneously. Furthermore, we showed

that the enhanced darkfield imaging based technique could successfully be ap-

plied for semi-quantification of cellular uptake of NPs, facilitated by highly specific

spectral profiles of NPs. In addition, when combined with other biological assays,

e.g. toxicity or oxidative stress measurements, HSI-M could contribute to eluci-

dating the mechanisms of biological effects of NPs by providing the information

on subcellular localization of NPs, their aggregation pattern and effects of ES on

NP suspendability/agglomeration (see SI for Correlations of Nanoparticle Uptake

as Quantified with HSI-M with Sub-toxic Effects of Nanoparticles and Figure

S3.11). The importance of the extracellular substances was clearly demonstrated

in this study – utilizing reference spectra obtained from ES coated NPs consider-

ably improved the quality of hyperspectral analysis of intracellular NPs most likely

due to the similar interactions taking place inside the cell. Semi-quantification of

uptake of NPs was significantly influenced by the type, size and concentration of

NPs as well as the time of exposing T. thermophila to NPs. In addition, the physi-

ology and phagocytosis rate of the protozoa varied between cells in the same

sample as well as between different exposures representing a highly variable

biological sample. In case of such systems, low sample throughput and long

analysis time still hamper the acquisition of statistically strong datasets. Further

work is encouraged to automate certain steps of the analysis, e.g. image acquisi-

tion and data treatment.


The study was supported by a Sciex-NMSch fellowship to Monika Mortimer, fund-

ing from the Swiss National Research Program NRP 64 "Opportunities and Risks

of Nanomaterials” to Vera Slaveykova, Thomas Bucheli and Alexander Gogos,

and Estonian Science Foundation (ETF8561 and IUT 23-5) to Anne Kahru. Nora

Bartolomé was funded by a Leonardo da Vinci scholarship. We also thank the

Bundesamt für Umwelt (BAFU) for financial support. Jim Beach (Cytoviva) is


107

gratefully acknowledged for helpful discussions and ideas on the HSI-M proce-

dure and Isabel Hilber for statistical data analysis.

Chapter 3

108


3.6.1 Experimental

3.6.1.1 Detailed description of image processing and analysis with hyperspectral

imaging microscopy (HSI-M)

a. Acquisition of reference and sample HSIs

Firstly, two different HSIs of the pure NPs under scrutiny were recorded (refer-

ence HSIs): one with the pure NPs and one where the NPs were pre-incubated

with protozoan extracellular substances (ES). Then, for the semi-quantification

analysis, 10 single cell images were collected randomly for each type of NP ex-

posure (sample HSIs). All HSIs in one batch were recorded at the same exposure

time. However, different exposure times were needed depending on the type of

the NPs.

b. Spectral smoothing

To reduce noise and smooth the shape of spectral curves, the CytoViva-feature

“adjacent band averaging” was used. Sample and reference HSIs were smoothed

at the same smoothing width of 5 nm.

c. Spectral correction

A lamp spectrum was recorded over a glass slide under the same conditions as

the sample slides (same glass slide type with Mowiol® 4-88 layer covered with a

glass slip) to perform a spectral correction of the obtained HSIs for the uneven

spectral response of the microscope optics, using the CytoViva Analysis feature

„Normalize for Lamp Spectrum“ inside ENVI.

d. Extraction of representative spectral libraries (SL) from the reference sam-

ple

The CytoViva-feature Particle Filter Analysis (PFA) was used to obtain repre-

sentative SL of NPs from the reference HSIs. This analysis was possible when

the reference sample was homogeneous and consisted of single spherical parti-

cles as was the case for Ag, Au and TiO2 NPs used in this study. Briefly, PFA

allows detection of particles or circularly-shaped objects in an image. The follow-

ing parameters were applied to achieve best quality of the obtained spectra: min-

imum signal intensity 1,000, maximum signal intensity 4,050 and a size threshold


109

of 9 connected pixels. After this, the different spectra of particles selected by PFA

were averaged per particle to create the final SL. The average spectral profiles,

which constituted a SL for a given NP, of each type of NP suspension were ob-

tained based on 536, 431, and 361 single averaged NP spectra for Ag, Au and

TiO2 NPs, respectively. This library was then used for further spectral angle map-

per (SAM) analysis. In the specific cases of CuO NPs and QDs it was not possi-

ble to apply PFA, because the reference NPs did not meet the above mentioned

conditions. Therefore, for these two types of NPs, 20 spectral profiles of different

particles from the reference material were collected manually and integrated into

an SL. In general, for every SL acquisition, selected endmembers were checked

not to contain spectra with (i) too low intensity (background noise), (ii) too high

intensity (peak shape deformation) and (iii) low number of connected pixels with

inhomogeneous spectral profiles.

e. Application of a spectral classification method to localize and semi-quantify

NPs in sample image

The SL obtained by PFA was then used with a classification algorithm (SAM) to

estimate the relative abundance of NPs per cell. SAM calculates the similarity

between the spectral profile in each pixel of the sample HSI and the SL obtained

from the reference HSI at each of the n wavelength bands. This is done by first

determining a vector in n dimensions that represents the distance from the origin

(dark) to the light intensity recorded in each band of the sample spectrum. The

direction of this vector in n-dimensional space is used to define a unit vector rep-

resenting the sample spectrum. The same procedure is performed for the refer-

ence spectrum. SAM then determines the angle between the two unit vectors.

The best spectral match occurs when the angle between these vectors is the

smallest. The most restrictive match was achieved when the angle between ref-

erence and sample spectra was ≤ 0.1 radians. Once the sample HSI was loaded

for SAM analysis a spectral subset was used in order to eliminate possible inter-

ference due to background noise. For each type of the NPs a specific spectral

subset was used (10-nm sized Ag NPs: 450 to 700 nm; 20-nm sized Ag NPs: 580

to 800 nm; 10-nm sized Au NPs: 550 to 800 nm; 20-nm sized Au NPs: 600 to

800 nm; TiO2 NPs: 400 to 800 nm; CuO NPs: 600 to 1,000 nm and QDs: 620 to

720 nm), depending on their main peak regions (Figure 2). In the final step of

SAM a classification image was created and overlaid onto the original sample

Chapter 3

110

HSI. Thereby, pixels that were classified to contain a spectral profile similar to the

reference SL were located and highlighted in the sample HSI. In general, as the

reference SL contained multiple endmembers (see Figure 1), the final classifica-

tion pixels in the image were composed of multiple colors, correlating with each

endmember. These were then combined by simply merging the different colors to

a single color, leading to a single matching result. For quality assurance, every

reference SL was used on HSIs of control cells that were not incubated with NPs.

No false positives were detected, except for TiO2 NPs.

3.6.1.2 Brightfield imaging

As a complementary method to HSI-M for uptake monitoring, samples were also

analyzed using a Leitz Laborlux S brightfield (BF) microscope equipped with a

Leica DFC290 camera at 100x magnification. This was possible, as NPs formed

agglomerates in the protozoans during incubation that were visible with this type

of microscopy (see Figure S3 for exemplary BF images). One-cell images were

saved as grayscale images and then analyzed in ImageJ (rsb.info.nih.gov/ij/ ) us-

ing the following steps: (i) thresholding, (ii) inverting image, (iii) particle analysis.

For all BF images, the same settings were applied. NP agglomerates appeared

as dark areas in the protozoa. The results were presented as agglomerate-

equivalent area per cell.

3.6.1.3 Viability assay

T. thermophila in 10 mM HEPES buffer, pH 7.0, was exposed to NPs at indicated

concentrations for 2 or 24 h in 5-mL tubes (Falcon) in a volume of 1 mL at cell

concentration of 5×105 cells mL-1. For detection of dead and membrane dam-

aged cells SYTOX® Green Nucleic Acid Stain (Life Technologies Europe B.V.,

Carlsberg, USA) was added to NP-exposed cells at a concentration of 1 µM and

incubated at room temperature for 10 min. The fluorescence of the DNA bound

stain was measured with flow cytometry. Protozoan suspension not exposed to

NPs was used as a negative control and protozoa exposed to 5 mg L-1 of Ag2SO4


111

(100% lethal concentration) and 1 mg Cu/L of CuSO4 (50% lethal concentration)

as a positive control [294, 315].

3.6.1.4 Oxidative stress assays

T. thermophila was exposed to NPs as in the viability assay (see above). For de-

tection of intracellular ROS and lipid oxidation CellROX® Green Reagent and

BODIPY® 581⁄591 C11 (Life Technologies Europe B.V., Carlsberg, USA), re-

spectively, were added to NP exposed cells at a concentration of 5 µM and incu-

bated at room temperature for 20 min. The fluorescence of the oxidized stain was

quantified by flow cytometry. Protozoan suspension not exposed to NPs was

used as a negative control and protozoa exposed to 100 µM cumene hydroperox-

ide (Life Technologies Europe B.V., Carlsberg, USA) as a positive control.

3.6.1.5 Flow cytometry

Accuri C6 Flow Cytometer® was used for fluorescence and light scattering

measurements in T. thermophila. In all cases 488 nm laser was used for excita-

tion and data of all the fluorescent markers used in this study were analyzed us-

ing the FL1 channel (emission at 530±15 nm).

Chapter 3

112

Figure S3.1: Hyperspectral images of reference NPs (see Table 3.1) incubated in protozoan

extracellular substances (ES). The images were acqu ired using 100x objective with a zoom

in of 6x.

Figure S3.2: Cytotoxicity induced by metal and meta l oxide NPs in Tetrahymena thermoph-

ila, quantified by measuring the fluorescence of the n onviable cell marker Sytox Green. All


113

the values were normalized to the fluorescence of p ositive control (protozoa exposed to

100% lethal concentration of Ag 2SO4). Nanoparticles were considered to cause lethality in

protozoa when the % of dead cells was above the res pective value in control (unexposed)

cells (above 5 % at 2 h and 5.5 % at 24 h). Thus, s ome cytotoxicity was observed only for

10 mg L -1 TiO2 NPs after 24-h exposure.

Figure S3.3: Bright field images of Tetrahymena thermophila exposed to NPs for 2 h (top

row) and 24 h (bottom row).

Figure S3.4: Darkfield images of a Tetrahymena thermophila cell exposed to 20 nm Ag na-

noparticles at 10 mg L -1 for 2 h. The images were taken at consecutive foca l planes along

Z-axis. Scale bar = 20 µm.

Figure S3.5: Hyperspectral images of Tetrahymena thermophila exposed to 20 nm Ag NPs

for 2 h. Red areas indicate matched pixels when spe ctral angle mapping (SAM) analysis

was performed on the same cell using spectral libra ries (SLs) from Ag NPs in aqueous

Chapter 3

114

suspensions (A) and SLs obtained from Ag NPs incuba ted in protozoan extracellular sub-

stances (B). Scale bar = 10 µm.

Table S3.1: Significance of reference NP incubation with ex-

tracellular substances, NP concentration and size a s well as

exposure time on the HSI analysis results (expresse d as %

matched pixels per cell area).

%Matched pixels Ag Au TiO 2 CuO

Extracellular sub s-

tances

*** n.s. n.s. n.s.

Concentration - *** *** ***

Size * ** - -

Exposure time *** *** n.s. ***

Significance levels: n.s. – not significant, *P<0.05, **P<0.01, ***P<0.001

Used model: yijkl=µ+αi+βj+γk+δe+εijkl

With yijkl: dependent variable (%matched pixels), µ: overall mean,

αi: extracellular substances y/n, βj: concentration (mg L-1), γk: size

(nm), δe: exposure time (h), εijkl: random error


115

Figure S3.6: Hyperspectral image (HSI) of a non-exp osed (control) Tetrahymena ther-

mophila cell analyzed by spectral angle mapping (SAM) with a spectral library (SL) of TiO 2.

Areas erroneously recognized as TiO 2 are highlighted in red and with an arrow. Scale ba r =

10 µm.

Figure S3.7: Spectral signature of CdSe/ZnS core/sh ell quantum dots.

Chapter 3

116

Figure S3.8: Exemplary spectral angle mapping (SAM) analysis performed on hyperspec-

tral images of (A) 10 nm Ag-incubated cells using s pectral libraries( SLs) of 10 nm Ag NP

(A1) and 20 nm Au NP (A2) reference particles and ( B) 20 nm Au-incubated cells using SLs

of 20 nm Au NP (B1) and 20 nm Ag NP (B2) reference particles. Pixels matched to the re-

spective reference library are indicated in red. Fa lse positives only occurred in case of B2,

where the spectra under scrutiny were very similar. Scale bar = 10 µm.


117

Figure S3.9: Bright field images of Tetrahymena thermophila exposed to Ag NPs (A, B) and

Au NPs (C, D) for 2 h (A, C) and 24 h (B, D). Image s were taken with a 40X objective.

Chapter 3

118

Figure S3.10: Comparison of hyperspectral imaging o f internalized metal NPs (% matched

pixels/cell area) with bright field image analysis (agglomerate area/cell area) for semi-

quantification of internalized nanoparticles. Data from the 2- and 24-h uptake experiments

with Au and Ag NPs were collectively correlated; se e Table 3.1 for concentrations of NPs

and Figure 5 for uptake data expressed as % matched pixels per cell area. a.u. – arbitrary

units.

3.6.1.6 Correlations of Nanoparticle Uptake as Quantified with HSI-M with Sub-

lethal Effects of Nanoparticles

As stated earlier, in the current study NP uptake experiments were conducted at

sub-lethal concentrations to T. thermophila (Figure S2). However, in some expo-

sure conditions NPs induced increased intracellular ROS and/or lipid peroxidation

in protozoa as measured by the fluorescence of CellROX® Green and BOD-

IPY®581⁄591 C11, respectively. We compared the rate of NP uptake by T. ther-

mophila with the ROS and lipid peroxidation induced by the NPs. Specifically, we

focused comparatively on inert Au NPs and particles which are known to dissolve

at a more extended rate, i.e., Ag NPs [316]. ROS generation and oxidative stress

are considered the best-accepted paradigm to assess and compare the toxicity of


119

NPs [317]. ROS generation by metal-based NPs can be triggered directly, pro-

moted by particle properties at the nanoscale, or indirectly by dissolved, toxic

metal ions from NPs. NPs can produce ROS at the nano-bio interface or inside

the cells, where acidic phagosomes or lysosomes may promote the dissolution of

NPs [317]. As lipid molecules constitute approximately 30–80% of biological

membranes by mass [318], peroxidation of lipids in response to ROS is a very

likely scenario and a good biomarker of cellular oxidative stress. Thus, consider-

ing the range of possible pathways for ROS generation and consequent reactions

in the cell, we expected the correlations between NP uptake and biological ef-

fects to depend on the NP type and dissolution properties. When we compared

the uptake versus ROS formation and lipid peroxidation induced by Au and Ag

NPs, in general, the correlations in case of Ag were better than these of Au NPs

(Figure S11). Ag NPs as more soluble particles could induce ROS proportionally

with the uptake into food vacuoles by releasing Ag ions from NPs in close proxim-

ity of biomolecules and membranes [319, 320]. Conversely, the uptake pattern of

Au NPs did not correlate well with the ROS generating potential of the particles.

The high values of fold change in fluorescence of CellROX Green when used in

the presence of 20nm Au NPs (Figure S11A) could partly be caused by the con-

tribution of surface plasmon resonance (SPR) of Au NPs to the fluorescence sig-

nal due to the overlapping wavelengths of absorbance/SPR of Au NPs (450/524

nm) and excitation/ emission of CellROX Green (485/520 nm). Interestingly, after

2 h of exposure, much higher Au NP uptake rate (expressed as percentage of

matched pixels per cell area) compared to Ag NP (9 times higher in case of

10 nm particles and 3 times higher in case of 20 nm NPs) resulted in the same

level of lipid peroxidation (Figure S11 C, D), indicating deviation from good corre-

lations between uptake and sub-lethal effects for the more inert NPs compared to

more dissolvable and bioreactive NPs.

Chapter 3

120

Figure S3.11: Scatter plots illustrating the relati onship between the potential of Au and Ag

nanoparticles at 10 mg L -1 to induce intracellular ROS and lipid peroxidation in Tetrahy-

mena thermophila and the cellular uptake levels of these NPs as det ermined by HSI. The

data labels in the graphs denote the size of the re spective NP/exposure time. a.u. –

arbitrary units.

121

Capabilities of asymmetric Chapter 4

flow field-flow fractionation coupled to

multi-angle light scattering to detect car-

bon nanotubes in soot and soil

Alexander Gogos, Ralf Kaegi, Renato Zenobi and Thomas D. Bucheli

Published in Environmental Science:Nano, 2014, 1(6): 584-594.

Reproduced by permission of the Royal Society of Chemistry.

Chapter 4

122

Abstract

Analytical detection and quantification of multi-walled carbon nanotubes

(MWCNTs) in complex matrices such as soils is very challenging. In an initial ap-

proach to this task, we identify MWCNTs by making use of their different (e.g.,

rod-like) shape compared to other (native) soil particles and in particular soot,

which is ubiquitously present in soils. A shape factor ρ, determined using asym-

metrical flow field-flow fractionation coupled to multi-angle light scattering (aF4-

MALS), was used to discriminate MWCNTs of different aspect ratios, as well as

mixtures of soot and MWCNTs, in pure suspensions. MALS results were addi-

tionally confirmed using automated electron microscopy image analysis. We then

analyzed different soil types which consistently showed ρ-values that differed

from pure MWCNTs. To test the performance of the method for MWCNT detec-

tion in such complex matrices, we conducted standard additions of a MWCNT as

well as soot to an agricultural soil. Extracts from these MWCNT-spiked soils

showed increased ρ-values compared to soot-spiked or native soil. The method

detection limit for the MWCNT was 1.6 to 4.0 mg g-1 soil and lies within the range

of commonly used black carbon quantification methods, but is much higher than

any currently predicted environmental concentration. Additionally, the method is

currently limited by a relatively narrow dynamic range of ρ. Despite these limita-

tions, our results suggest that aF4-MALS provides specific shape information that

may be linked to an actual MWCNT presence in soils. Further method improve-

ment potential is outlined along different steps of the workflow.

Capabilities of aF4 coupled to MALS to detect carbon nanotubes in soot and soil

123

4.1 Introduction

Since Iijima first described tube-like structures of sp2-hybridized carbon [69] - now

known as carbon nanotubes (CNTs) - this new material type has undergone an

enormous development, which is reflected by a production volume that already

exceeds several thousand tons per year [24]. CNTs are increasingly used in con-

sumer products [24, 321] that include, e.g., lightweight composites in cars and

aircrafts, antifouling paints, batteries and water filters. As the prices of multi-

walled carbon nanotubes (MWCNTs) are still orders of magnitude lower than for

the single-wall types (SWCNTs) [24], they prevail in terms of production volumes.

During their life cycle, these materials could potentially be released into the envi-

ronment via different routes [26, 234, 322]. They might even be intentionally ap-

plied in agriculture, as several beneficial effects of CNTs on plants were postulat-

ed, such as enhanced water uptake, seed germination, and cell growth [237,

323]. Such direct applications would induce much higher fluxes into soils than

predicted to date [324]. In contrast, some studies suggest that CNTs negatively

affect soil microbial communities as well as plants [248, 325, 326], after being

released into the environment. This highlights the need to efficiently assess the

environmental distribution of CNTs. However, detection and quantification meth-

ods for CNTs in soils and sediments are still not well established. One of the

main challenges in CNT quantification in soils and sediments is the ubiquitous

presence of black carbon (BC) particles. These originate from wildfires or incom-

plete combustion of fossil fuels [95], and have chemical and physical properties

similar to CNTs. First attempts to quantify CNTs in soils were made by Sobek

and Bucheli in 2009 [94], using standard additions and a chemothermal oxidation

method (CTO-375). Unfortunately, this method alone is not able to differentiate

between soot-BC and CNT-BC. Alternative methods, such as thermal/optical

transmittance/reflectance [97] or thermogravimetry (coupled to mass spectrome-

try [96]), make use of the lower thermal stability of some SWCNT compared to

soot. However, the thermal stability may vary depending on the length and the

thickness of the CNTs, and even be comparable to soot [94], resulting in potential

co-isolation. Also, quantification of CNTs may be influenced by matrix effects in

more complex samples, such as urban dust and sediments [96, 97]. Further

promising techniques include extraction of CNTs with suitable surfactants and

Chapter 4

124

near-infrared fluorescence (NIRF) [102], or microwave-induced heating [98].

However, the former is only applicable to SWCNT, and the latter lacks specificity.

Given the limitations of these different methods, it is evident that quantification of

CNTs in real-world samples will probably require applying a suite of complemen-

tary methods in one single workflow. A parameter that is very special for CNTs is

shape: CNTs are rod-like to randomly coiled structures with high aspect ratios,

whereas soot particles consist of spherical primary particles [92] aggregated to

fractal-like structures [327]. A technique that has already been employed for

shape description of environmental colloids [113] is asymmetric flow field-flow

fractionation (aF4) coupled to multi-angle light scattering (MALS). Asymmetric

flow field-flow fractionation in combination with UV-Vis or electron microscopy

(EM) has previously been used to separate CNTs in suspension [328-330], and

coupled with MALS for measuring their length and dispersion state [331, 332].

Thus, aF4-MALS may be able to detect changes in the overall shape of a natural

sample once CNTs are present.

Here, we describe the development of an aF4-MALS method to differentiate be-

tween a MWCNT and soot in synthetic mixtures of these materials. We used au-

tomated EM in combination with image analysis tools to confirm the aF4-MALS

results, and tested the robustness and performance of aF4-MALS to identify a

MWCNT in natural soil samples spiked with this analyte. The potential and limita-

tions of this approach are discussed in terms of methodological figures of merit,

and its applicability to actual environmental samples.

4.1.1 Theory

The principles and theory of aF4 and MALS have been presented in detail else-

where [128, 333]. Here we will briefly describe the main concepts of MALS, be-

cause our data is mainly based on this technique. All MALS measurements are

commonly evaluated using equation (4.1):

3(4)56 = MP(θ) (4.1)

where R(θ) is the Rayleigh ratio, c the concentration of the sample, K a contrast

factor, M the molecular weight of the sample, θ the scattering angle, and P(θ) the

particle scattering function. The latter describes the angular dependence of the


125

scattered light and is dependent on the size and shape of the particles. Based on

the MALS data and equation (4.1), the radius of gyration (rg) can be extrapolated

as the slope of the function 3(4)56 vs. sin=(θ/2)at zero angle. The radius of gyra-

tion can be interpreted as the weighted average of all possible radii of a particle

from its center of mass. Therefore, rg is susceptible to changes in particle shape.

The hydrodynamic radius (rh) is approximated for non-spherical particles as the

radius of a sphere with the same diffusion behavior. It can be obtained either

from aF4 retention time calibrated with latex spheres with a known size or by

online dynamic light scattering (DLS) measurements. The ratio of these two radii

gives a shape factor

ϱ = ABAC (4.2)

that is a direct measure for particle shape [334]. Homogenous spheres show a ρ

of 0.775 [113], and values >0.775 indicate a deviation from a spherical particle

shape. In addition, ρ is proportional to a particle’s aspect ratio [334], thus for high

aspect ratio–particles, such as MWCNTs, high values for ρ can be expected.

4.2 Materials and methods

4.2.1 Chemicals and analytes

Ammonium nitrate (NH4NO3, ≥99%), sodium azide (NaN3, ≥99.5%) and sodium

deoxycholate (SDC, ≥98%) were purchased from Sigma-Aldrich (Buchs, Swit-

zerland). Nanosphere™ size standards were purchased from Thermo Scientific

(Fremont, CA). Carrier solutions for aF4 were prepared in Milli-Q water (Millipore,

Zug, Switzerland). Before use, all carrier solutions were filtered through 0.1 µm

Durapore filters (Millipore). Long pristine (MW1), long carboxylized (MW2), short

pristine (MW3) and short carboxylized (MW4) MWCNTs were purchased from

Cheap Tubes Inc. (Brattleboro, VT). All MWCNTs were declared to be of 90–95%

purity and to have an elemental carbon content of >96%. The MWCNTs were

used as received without further purification. However, characterization of the

very same batch of MWCNTs has been carried out over several years in our

Chapter 4

126

group [94, 335, 336], and detailed data are provided in the electronic supporting

information (Table S4.1). Here, we additionally imaged the suspensions of the

MWCNTs using scanning electron microscopy (SEM) (see Figure S4.1). Forklift

Diesel soot (Standard Reference Material (SRM) 2975, National Institute of

Standards and Technology (NIST), Gaithersburg, MD) with spherical primary par-

ticles in the range of 35 nm [92] served as a BC standard. For reasons of sim-

plicity, we here refer to this material as “soot”.

4.2.2 Soils and sediments

A series of well-characterized soils and sediments served as natural samples for

our method, i.e., a loamy sand agricultural soil from Germany (LUFA Standard

soil 2.2., Landwirtschaftliche Untersuchungs- und Forschungsanstalt, Speyer,

Germany), a clay (vertisol) soil from Australia, several soils from the Swiss Soil

Monitoring Network (NABO) and a marine sediment from Baltimore Harbor, USA

(NIST SRM 1941b). The clay soil as well as the sediment were part of an earlier

BC inter-laboratory ring trial, and thus well characterized [337]. The NABO soils

had been analyzed for BC content by Agarwal and Bucheli [93]. For detailed pa-

rameters and description of the used soils see Table S4.2.

4.2.3 Analyte suspensions and sample extracts

4.2.3.1 Pure supensions and corresponding mixtures of MW1 and soot

For aF4-MALS measurements, MWCNTs and soot were dispersed in 2%

SDC/0.05% NaN3 solution at a nominal concentration of 50 mg L-1 and sonicated

for 3x10 min in a 720W high-power sonication bath (Bandelin, Berlin, Germany)

with intermediate shaking. Sodium azide was added to both samples and aF4

carrier solution to prevent bacterial growth. Subsequently, the dispersion was

centrifuged at 17’500g for 10 min (Sorvall Superspeed, DuPont Instruments,

Newton, CT) to remove agglomerates. The overlaying half of the volume then

served as a working suspension. This methodology has been shown before to

yield very stable and well dispersed suspensions [102]. Suspensions were free of

aggregates (see Fig. S1 for SEM images of the suspensions) and very stable,

with ζ-potentials around -50 mV. Residual MWCNT and soot concentrations in

the supernatants were determined by UV-Vis (at λ=289nm), calibrated with differ-

ent concentrations (1-15 µg mL-1) of the analytes in non-centrifuged dispersions.


127

Appropriate volumes of MW1 and soot suspensions were then mixed to yield

mass ratios of 0.8:1, 1:1, 1.3:1, 2:1 and 4:1 (corresponding to MW1 mass frac-

tions of 44.5, 50.0, 56.5, 66.7 and 80%).

4.2.3.2 Extractions of soils and sediments

Prior to use, each soil was dried at 105°C until constant weight, sieved (<2 mm

mesh gauge) and ground using a ball mill (Retsch MM400, Haan, Germany).

Then, 30 mg were extracted in 10 mL 2% SDC/0.05% NaN3 solution as de-

scribed above for pure suspensions. Each sample was extracted in 3 replicates.

4.2.3.3 Standard additions of MW1 and soot to soil extracts

For simulation of 100% analyte extraction efficiency and determination of a nomi-

nal method detection limit (MDL), standard additions of MW1 and soot of 5, 13,

25 and 50 µg mL-1 (corresponding to 1.6, 4.0, 8.4 and 16.4 mg g-1 of soil, respec-

tively) were performed to extracts of the Lufa 2.2 agricultural soil. These soil ex-

tracts were obtained by extraction of 120 mg of unspiked soil with 40 mL 2%

SDC/0.05% NaN3 as described above for pure suspensions, including the cen-

trifugation step. One 20-mL aliquot was then spiked with the respective

MW1/soot powder to a concentration of 100 µg mL-1, sonicated as described be-

fore, and then diluted into the unspiked aliquot, to maintain the original matrix. At

this stage, another centrifugation step was omitted to ensure that the nominal

analyte concentrations were still met. Therefore, in these samples, agglomerates

were still present.

4.2.3.4 Standard additions of MW1 and soot to soil

Both MW1 and soot were also spiked directly to the Lufa 2.2. agricultural soil at

concentrations identical to those of the extracts described before (1.6, 4.0, 8.4

and 16.4 mg g-1 of soil) to determine an effective MDL. Spiking was performed by

mixing the respective mass of the analyte into the soil with a spatula and carefully

disintegrating visible agglomerates/chunks. The spiked soil was then placed into

a Turbula powder mixer (Turbula System Schatz, Willy A. Bachofen AG, Switzer-

land) for 24h. From the resulting samples, 30 mg were extracted in 10 mL 2%

SDC/0.05% NaN3 solution according to the method described above for pure

suspensions and unspiked soils, including all steps. Each sample was extracted

in 3 replicates and then joined to form a collective sample.

Chapter 4

128

4.2.3.5 Separation (aF4) and online detection (UV-Vis, MALS)

For aF4, we used a commercial apparatus (AF2000, Postnova Analytics GmbH,

Landsberg, Germany). This instrument was connected online to a UV-Vis diode

array detector (PN3241, Postnova Analytics), a 21-angle MALS detector

(PN3621, Postnova Analytics), and a Malvern Zetasizer (Infors, Bottmingen,

Switzerland). The aF4 channel was trapezoidal-shaped, 350 µm thick (defined by

a spacer), 29.8 cm long (inlet to outlet), and had a 2 cm maximum width. All runs

were done with a 10-5 M NH4NO3/0.02% NaN3 mobile phase (pH 6.5), which has

previously been shown to be suitable for CNT fractionation in aF4[332]. Particle

recoveries over aF4 were tested with several commonly used membranes (see

Table S4.3), and were highly variable for both CNTs and soot. Since the objective

of this paper is a differentiation of particles based on shape, we selected regen-

erated cellulose membranes (RC; 10kDa). This membrane type was the only one

that provided equal recoveries (~50%) for both types of particles. Operating con-

ditions of the aF4 were optimized to separate both size standards, as well as the

samples with the same method. Membranes were changed at the latest after 20

to 25 injections. All injections were performed in triplicate with an autosampler

(PN5300, Postnova Analytics). Further fractionation conditions were as follows:

injection volume adjusted to 5 µg absolute analyte mass (280-410 µL, pure sus-

pensions) or 500 µL (natural samples and standard addition experiments), detec-

tor flow 0.5 ml min-1, cross flow 1 ml min-1 with a power gradient of 0.2, injection

flow 0.2 ml min-1, injection time 10 min and fractionation time 40 min. At the end

of the fractionation process, a 7-min rinse step without cross-flow but full detector

flow was applied, to minimize the risk of cross-contamination between injections.

4.2.3.6 Fraction collection, automated electron microscopy and image analysis

Suspensions of pure MW1, soot, and mixtures thereof were fractionated after aF4

using a fraction collector (PN8050, Postnova Analytics). For their analysis by

SEM, 375 µL of the collected fractions were filled into a conical Eppendorf tube

with a carbon-coated TEM grid on a plastic stopper. This setup resulted in a wa-

ter column of 5 mm height above the TEM grid. The particles were deposited on

the TEM grids by centrifugation (1h at 16’000 g) using a swinging bucket rotor.

Images were recorded on a SEM (Nova Nano-SEM 230, FEI, USA) with a trans-

mission electron detector. The bright field signal was used for image formation. In


129

total, 40 (soot), 30 (MW1) and 20 (mixture) images were automatically recorded

at a fixed magnification (20’000) using INCA suite 4.15 (Oxford Instruments, Ox-

ford, UK). All images were automatically processed in Fiji (background removal

and thresholding) to obtain binary images and then skeletonized using the ‘Skele-

tonize 2D/3D’ and ‘Analyze Skeleton 2D/3D’ plugins [338-340]. The “longest-

shortest” branch was then used as the skeleton length. Distributions of this skele-

ton length from the pure materials then served to calculate percentages of MW1

and soot in mixed suspensions by applying the following linear model [341]:

XEFG = α ∙ XJKLMN + β ∙ XQRRS + ε (4.3)

where X is the skeleton length distribution of the respective material (see index),

α and β the relative contribution of the corresponding material to the total number

based amount of particles, and ε accounts for the difference between modeled

and observed distributions and for random errors.

4.2.3.7 MALS data analysis

Data analysis was always performed on the last of three replicate injections (if not

stated otherwise), to ensure data acquisition at stable run conditions. MALS data

were analyzed with NovaMALS 1.0.0.9. (Postnova Analytics). Since the actual

geometry of the particles (MWCNT and soot) was unknown, a 5th grade polyno-

mial Debye fit for P(θ) that implies an arbitrary shape[113] was used for rg calcu-

lations (for examples, see Figure S4.2). The rh was determined by calibrating the

retention time (UV-Vis) with two different mixtures of Nanosphere™ size stand-

ards (10, 50, 100, 175 nm and 30, 75, 150, 200 nm, Figure S4.3). The resulting

calibrated values for rh were in good agreement with values from online DLS

measurements (7-19% deviation). According to equation (2), ρ was then calculat-

ed over the aF4 retention time from rg and rh values, and represented in a frac-

togram. For further evaluation, a shape factor difference

∆ρF = WX,Y�WZB,YWZB,Y

(4)

was calculated for each point of the retention time i, where ρc,i is the shape factor

of a given analyte concentration and ρbg,i is the shape factor of the native back-

ground (e.g., soot, soil) at the same time point. Values for ∆ρF were then aver-

aged over 50% of the MALS 92° peak width for each concentration.

Chapter 4

130

4.2.3.8 Quality control, method validation and statistics

A correct normalization of the light scattering detector was assured by repeated

measurements of size standards, and subsequent verification of obtained rg val-

ues. Blank measurements were run regularly to assure artefact free analyses.

Additionally, alternate injections of MW1 and soot on the same membrane were

performed to ensure that no carry-over effects occurred.

The recovery from aF4 runs was determined as follows:

R(%) = ]]" ∙ 100 (4.5)

where S and S0 are the peak areas without void peak of the detector signal (UV,

λ=289 nm/MALS 92°), obtained with and without cross-flow, respectively. Addi-

tionally, we compared MALS 92° peak areas of samples that were spiked to soil

with those that were spiked to soil extracts (simulated 100% extraction efficiency)

to indirectly calculate analyte extraction efficiencies. Reproducibility of ρ over re-

tention time was determined by three independent measurements and monitored

by repeated measurements of known samples (e.g., the Lufa 2.2 soil extract).

To determine the instrumental limit of detection (LOD), we measured ρ distribu-

tions of decreasing MW1 amounts (from 5 down to 0.05 µg, constant injection

volume) in pure suspensions. Here, we define the LOD to be the lowest injected

MW1 mass, until which the distribution of ρ remains stable.

Method detection limits were calculated for the matrices soot and soil. Three dif-

ferent cases were evaluated (for the description of the samples see sections be-

fore): (i) mixtures of MW1 and soot (MDLsoot), (ii) standard additions of MW1 to

soil extracts (nominal MDLsoil) and (iii) standard additions of MW1 to soil (effective

MDLsoil). For the first one, a t-test was applied on the the ρ-values obtained over

50% of the MALS 92° peak width to compare each mixture (one replicate, n=1)

with pure soot. For the second and third, the area under each ρ-distribution (three

independent measurements, n=3) was calculated as “area under curve”, evaluat-

ed using one-way analysis of variance (ANOVA) and finally compared by pair-

wise comparison with a Bonferroni test. To compare these results with a more


131

conventional way of determining MDL, we used the well-established concept by

Keith et al. [342], according to which the MDL is defined as three standard devia-

tions of a blank sample (3σ), the detection criterion being (St-Sb) > 3σ (with St

representing the average signal for the sample and Sb the average signal for the

blank). All statistical analyses were done with the software R (version 3.01, the R

Foundation for Statistical Computing) integrated in RStudio (version 0.97.551,

RStudio, Boston, MA).


4.3.1 General features of aF4-MALS fractograms and consequences for their

interpretation

During all measurements, rg values decreased with increasing time in the begin-

ning of the aF4 fractograms (Figure S4.2). This was caused by void/steric elution

of particles (≤160 nm rg) that were not retained after focusing. After this

void/steric elution, particles always showed increasing rg values with increasing

time (normal mode) throughout the remaining elution. Because calibrated rh val-

ues referred to normal mode elution (see also Figure S4.3), meaningful ρ-values

could only be obtained beyond this inversion point. This also explains the high ρ

values in the beginning of the fractograms (Figure 4.1) which were therefore dis-

regarded. Toward the end of the peak (e.g., after 40 to 45 min, Figure S4.2), rg

values sometimes showed a slower increase compared to rh, which led to a

steep decrease of ρ, depending on the particle or respective mixture. This was

due to a more pronounced curvature of 3(4)56 at larger particle sizes that could not

be fitted correctly anymore using the Debye algorithm. Consequently, for further

data evaluation, we chose 50% of the MALS 92° peak width (full width at half

maximum), where both effects were negligible.

An injected mass of 0.5 µg pure MW1 was the lowest amount until which the ρ

distribution remained stable (Figure S4.4), and was therefore defined as LOD.

While the MALS 92° signal-to-noise ratio (S/N) was still above 10 (Figure S4.4),

we generally preferred ρ over the more conventional S/N ratio at a given (shape

dependent) MALS angle as a measure of detection limit, because it provides the

Chapter 4

132

more integral analyte signal information. Moreover, we chose injection amounts

>2.5 µg as a starting point for all subsequent experiments to measure in a range

well above the LOD.

Figure 4.1: (A) Fractograms obtained by aF4-MALS wi th shape factor ρ (filled circles) for

equal mass injections (5 µg) of different types of pure MWCNT suspensions: MW1 (pris-

tine, medium aspect ratio (1500), ●), MW2 (-COOH, high aspect ratio (3750), ●), MW3 (pris-

tine, low aspect ratio (200), ●) and MW4 (-COOH, low aspect ratio (200), ●). Vertical lines

indicate the transition points between void/steric and normal mode elution and solid lines

show the 92° MALS signal for each type of MWCNT in the respective color. The shaded

area corresponds to particle sizes too large for ap propriate fitting using the Debye algo-

rithm. (B) Average shape factor ρ obtained from 50% of the MALS 92° peak width vs. n omi-

nal aspect ratio of the different MWCNTs. Bars repr esent the standard deviation of the ρ-

values over the selected retention time window.

4.3.2 aF4-MALS analysis of MWCNTs and MW1-soot mixtures

To assess the relation between ρ and nominal aspect ratios, we measured vari-

ous MWCNTs that differed in this parameter (Figure 4.1). Shape factors for these

MWCNTs were found to correlate with their nominal aspect ratios (R2=0.877,

y=6·10-5x+1.052, p=0.058, Figure 4.1B). The dynamic range, however, was less

pronounced than expected. This may be due to selective losses of longer

MWCNT to the aF4 membrane (as described in detail in the following section).

Consequently, recoveries for MWCNTs (and soot) in aF4 were highly variable,

depending also on the nature of the membrane material (see Table S4.3), as well

as on carrier composition [332] and cross-flow selection [124]. After precondition-


133

ing a 10 kDa RC membrane with two injections, we obtained stable recoveries of

50 ± 9% for soot and 50 ± 5% for the pristine long MWCNT (MW1, also see Table

S4.1 and S4.3) determined by aF4-UV/Vis analysis. These recoveries were fa-

vorable for our type of analysis, as they were similar for both kinds of particles,

avoiding a bias on the shape factors from the start. These values were, however,

significantly lower compared to those reported by Gigault et al. [332] for short

carboxylized MWCNTs (94 ± 2%) under similar separation conditions. We meas-

ured such shorter, functionalized MWCNTs as well (see Figure 4.1), and

achieved recoveries of around 75% in the first injections, which then stabilized

between 41 and 49%, possibly indicating interactions between mobile phase-

dispersed and membrane-bound MWCNTs.

Figure 4.2A shows the ρ-value fractograms for soot, MW1, and mixtures of both.

Soot exhibited rather uniform ρ-values in the retention time window between 23

and 37 min. With an increasing fraction of MW1 in the mixtures, the elution profile

of the ρ-values developed an increasing maximum at about 31 min, similar to

pure MW1. The increase of this maximum was concentration dependent (Figure

4.2B). In addition, the increase in ρ was different at different time points (Figure

4.2A). Such an increase in ρ points to a separation by length, which has been

described before using EM [328, 329]. However, fractions of MW1 in the

MW1/soot mixtures remained rather constant throughout the fractogram (see

Figure S4.5). Thus, it was not possible to physically separate MW1 and soot us-

ing aF4 under the selected conditions.

The increase in the MW1 fraction in the mixtures correlated well with the average

∆ρ of each concentration (R2=0.997, y=0.0041x-0.146, p=0.0025, Figure 4.2B).

Statistical analysis (t-test, p<0.001) resulted in a MDLsoot of 44.4% (w/w, corre-

sponding to 2.2 µg injected MW1 mass). Conversely, a soot content of 20% was

significantly (t-test, p<0.001) discriminated from MW1. Amorphous carbon/soot

contents in commercial CNT products can range up to approx. 50% [343], [344],

and would thus influence the shape properties of the overall product. Thus, the

method presented here may contribute to identifying the soot content of CNT

powders after production.

Chapter 4

134

Figure 4.2: (A) Fractograms obtained by aF4-MALS wi th shape factor ρ (symbols) for equal

mass injections (5 µg) of a pristine long MWCNT (MW 1, ●), soot ( ●) and different mixtures

(w/w) of both: 44.5% ( ♦), 50.0% (■), 56.5% (■), 66.7% (▲), and 80.0% (▼) MW1 in soot. The

vertical line indicates the average transition poin t between void/steric and normal mode

elution and solid lines show the 92° MALS signal of pure soot (black) and pure MW1 (red).

The shaded area corresponds to particle sizes too l arge for appropriate fitting using the

Debye algorithm. (B) Average shape factor differenc e ∆ρ obtained from 50% of the MALS

92° peak width vs. MWCNT content in the mixtures. B ars represent the standard deviation

of the ∆ρ-values over the selected retention time window.

4.3.3 Orthogonal confirmation of aF4-MALS results using automated EM in

combination with image analysis

Fractions of MW1, soot and a mixture of both (1:1 w/w), collected from aF4 anal-

yses similar to those displayed in Figure 4.2, were imaged and analyzed using

automated EM in combination with image analysis tools (Figure 4.3). The skele-

ton length distribution of MW1 and soot was different and allowed a quantification

of their fractions in a mixed sample after aF4. This was accomplished by a linear

combination fit using the skeleton length distributions of pure MW1 and soot as

references. Results indicated that the mixture consisted of 34% MW1 and 58%

soot. The linear model explained 92% of the observed skeleton length distribution

of the mixture, while 8% remained unexplained. The fraction of MW1 derived

from the EM analysis was somewhat lower than the corresponding fraction de-

termined from aF4-MALS generated ρ-values averaged over the retention time

collected for EM analysis (52% MW1, 48% soot). Deviations between the two

methods may be explained by artifacts related to the sample preparation protocol


135

for EM analysis, resulting in different image qualities that affected the results from

the image analysis procedure to various degrees. Common EM artifacts, such as

particle recognition, overlapping particles or contaminations on the EM grid may

affect the length distributions of pure MWCNT and soot source profiles and thus

the particle quantification. The observed contamination effects were mostly relat-

ed to carbonaceous materials condensing at the location of the electron beam

which resulted in a darkening and loss of contrast of the images. An additional

cleanup, such as a thermal treatment under vacuum, may reduce these contami-

nation effects and thus help to increase image qualities and the figures of merit of

the EM technique. If number based losses in the aF4 can additionally be quanti-

fied, the combination of aF4 - as a technique able to gently reduce sample com-

plexity - with EM analysis holds the potential to obtain number based concentra-

tions of MWCNTs and soot from complex samples.

Furthermore, EM analysis of aF4 fractions revealed additional information on the

behavior of MW1 in the aF4 channel. Nanotube-shaped particles were modeled

to elute in normal mode up to 500 nm rod length (1 nm diameter) [345]. Nominal

lengths of MW1, however, were much larger (see Table S4.1). Image analysis of

EM images indicated that eluting MW1 ranged from 0.1 to 1 µm in length, with a

maximum number at around 300 nm (Figure 4.3). These findings support the hy-

pothesis that longer MWCNTs initially present (see Figure S4.6) were lost, most

likely on the aF4 membrane, and that skeleton lengths of up to 1 µm eluted in

normal mode. As discussed in greater detail by Phelan and Bauer [345], the loss

of longer CNTs may be explained by decreased rotation with increasing length

and alignment of the particles at the accumulation wall.

Chapter 4

136

Figure 4.3: EM image analysis workflow (from left t o right: original, thresholded and skele-

tonized images with longest-shortest branch through the particle (red) and side branches

(grey)) and skeleton length distributions obtained from skeletonized images of MW1, soot

and a 1:1 (w/w) mixture of both. Images were taken and analyzed from aF4 fractions col-

lected during 28-32 min retention time. The distrib utions of MW1 and soot (solid/dashed,

red,: MW1; black: soot) are displayed together with the 1:1 (w/w) mixture (solid, yellow).

4.3.4 aF4-MALS analysis of native soils

In a next step, we measured ρ values of a diverse range of native soil and sedi-

ment samples to assess the range of possible background situations (Figure 4.4,

for raw data, see Figure S4.7). Recoveries of their extracts over aF4 ranged be-

tween 43% (NIST SRM 1941b) and 66% (BC Vertisol), determined using the

MALS 92° peak areas (n=3). Values for ρ of all analyzed soils/sediments were

different from pure MWCNTs (Fig 1B) and varied between approx. 0.65 (NABO

89) and 1.0 (BC Vertisol) (Figure 4.4). They should thus provide enough contrast

for detection of spiked MW1, with a ∆ρmin of ~0.26 (Lufa 2.2) and a ∆ρmax of ~0.61

(NABO 89). Colloids extracted from soil have been measured by von der Kam-

mer et al. [113], who obtained higher ρ-values (ρaverage =1.12) compared to

extracts of the native soils in this study. This difference may stem from different

sample preparation (e.g., lower centrifugation forces and different extraction con-


137

ditions used by von der Kammer et al. [113]) as well as chemical/physical differ-

ences between the analyzed soils. For example, organic soils such as NABO 67

and 89 showed lower ρ values than some of the clay soils (NABO 1 and 46) or

sandy soils such as the Lufa 2.2.

Figure 4.4: Compilation of average shape factor ρ values obtained from 50% of the MALS

92° peak width of suspensions of (i) pure MW1 and s oot (■), (ii) native soils ( ▼): Lufa 2.2

(agricultural soil), BC Vertisol (Clay vertisol fro m Hammes et al. [337] ringtrial), NABO 46

(agricultural soil), NABO 1 (grassland soil), NABO 89 (turf) and NABO 67 (organic soil from

a vegetable garden) and (iii) a marine sediment ( ▲): NIST SRM 1941b. Bars represent the

standard deviation of the ρ-values over the selected retention time window.

4.3.5 aF4-MALS analysis of MW1 and soot spiked to soil extracts

While differences in ρ between pure MW1 and soil reached values up to ∆ρ~0.61

(Figure 4.4), we tested whether such differences also apply to mixtures of both.

To do so, we selected the Lufa 2.2 agricultural soil for standard addition experi-

ments. This soil had several advantages; (i) it is a possible CNT recipient matrix

due to its agricultural origin, (ii) it is well characterized and commercially available

Chapter 4

138

to other researchers in its function as a reference soil and (iii) it provides a “worst

case” scenario in terms of ρ, as it showed one of the highest values among the

soils we studied, similar to pure soot (Figure 4.4).

Asymmetric flow field-flow fractionation served to eliminate some of the matrix

constituents that would otherwise disturb MALS detection and ρ determination.

Compared to pure MW1 suspensions (Figure 4.2A), higher ratios of the void peak

relative to the analytically accessible part of the MALS 92° peak were observed

(see Figure S4.8). Thus, some soil constituents that were not retained after fo-

cusing were thereby separated from the peak of interest.

Spiking of MW1 to soil extracts (Figure 4.5A, for raw data see Figure S4.8) led to

a concentration dependent increase in ∆ρ up to an apparent plateau above 25 µg

mL-1 (corresponding to 12.5 µg injected mass). The highest concentration

showed ρ-values comparable to pure MW1 (up to approx. 1.2; see Figure 4.2A,

Figure 4.4 and S4.8). Obviously, the soil extract did not influence the aF4 -MALS

analysis negatively and provided a contrasting background for the MWCNTs, as

all ∆ρ values were above the soil baseline (∆ρ=0). Conversely, spiking of soot

yielded negative ∆ρ-values (Figure 4.5B). Absolute ρ-values were comparable to

pure soot (approximately 0.9; see comparatively Figure 4.2A and S4.6). Also, ρ

values for soot were again more uniform compared to MWCNTs over the main

peak region (Figure S4.7). This behavior could already be observed in pure sus-

pensions (Figure 4.2A). Interestingly, the lowest soot concentration showed the

highest difference in ρ to the blank soil, while the highest soot concentration was

identical to it. Currently, we cannot provide a satisfying explanation for this obser-

vation.


139

Figure 4.5: Average shape factor differences ∆ρ obtained from 50% of the MALS 92° peak

width in relation to spiked concentrations and inje cted masses of (A) MW1 and (B) soot,

both spiked to soil extracts of the Lufa 2.2 soil t hat is represented by the dashed red line

(∆ρ=0). Bars indicate the standard deviation of the ρ-values over the selected retention

time window.

4.3.6 aF4-MALS analysis of MW1 and soot spiked to soils

When MW1 was added directly to soil and then extracted (Figure 4.6A, for raw

data see Figure S4.9), ∆ρ increased with increasing MW1 concentration, similar

to Figure 4.5A, but with a lower standard deviation. Shape factor differences also

plateaued between 8.4 and 16.4 mg g-1. The highest MW1 concentrations again

showed ρ-values comparable to pure MW1 (see comparatively Figure 4.2A and

S4.9). No concentration-dependent increase in ∆ρ was observed upon addition of

soot (Figure 6B), and ∆ρ-values for the different concentrations were consistently

below the soil baseline (∆ρ=0). This is surprising, as from Figure 4.4 and 4.5B, it

could have been expected that ∆ρ for soot mixed to the Lufa 2.2 soil should be

closer to, or even overlapping, the baseline (∆ρ=0). Although this is again not

easily explained, the different ∆ρ behavior of MW1 and soot in soil nicely shows

the different reaction of the measurement system to the presence of differently

shaped particles.

Chapter 4

140

Figure 4.6: Average shape factor differences ∆ρ obtained from 50% of the MALS 92° peak

width in relation to spiked concentrations of (A) M W1 and (B) soot, both spiked to the Lufa

2.2 soil that is represented by the dashed red line (∆ρ=0). Bars indicate the standard devia-

tion of the ρ-values over the selected retention time window.

4.3.7 Quality control and method validation

Sodium deoxycholate was employed far above the critical micelle concentration

(2.4 mM [346] or 0.09%, respectively) to efficiently extract MWCNTs. In this con-

centration it did not influence the measurements (see blank analyses in Figure

S4.10B). Alternating injections of pure MW1 and soot suspensions showed that

no detectable carry-over effects occurred between injections in terms of ρ-

development (Figure S4.10A). Between these injections, ρ-values showed rela-

tive standard deviations (RSDs) of approx. 7% (soot) to 8% (MW1). Reproducibil-

ity of aF4-MALS analyses of MW1 spiked to soil extracts and soils was highest in

the peak center and end; approx. 2-10% RSD in case of soil extracts (Figure

S4.11A) and 1-6% in case of soils (Figure S4.11B). The higher reproducibility of

the latter may be attributed to the centrifugation step, which removes agglomer-

ates.

When applied to the soil, the dynamic range of the measured signal ρ spread be-

tween 0.7 and 1.2. This range is rather narrow and limits the approach in general.

It could potentially be extended if losses of longer MWCNTs during aF4 separa-


141

tion are reduced. For example, Gigault et al. achieved recoveries of 89% for

SWCNTs and measured ρ-values of up to 3.6, corresponding to SWCNT lengths

up to 2 µm [124]. Also, depending on the type of the CNT, ρ may be higher (Fig-

ure 4.1B, MW2).

The nominal MDLsoil obtained from the standard additions to soil extracts was 5

µg mL-1 (corresponding to 2.5 µg injected mass or 1.6 mg g-1 of soil, respectively)

based on statistical analysis. This MDL was comparable to the MDLsoot in terms

of injected mass, which nicely illustrates the selectivity of the method in these

different matrices. Applying the approach by Keith et al. [342], we calculated the

nominal MDLsoil according to the criterion (St-Sb) > 3σ, with St = 0.956 (average

ρ-value of three replicate measurements of the lowest concentration within 50%

of the MALS 92° peak width), Sb = 0.868 (average ρ-value of the blank) and 3σ =

0.051 (three times the standard deviation of the blank). Thus, the lowest MWCNT

concentration fulfilled the criterion with 0.088>0.051 and both methods resulted in

the same nominal MDLsoil. This also held true for the effective MDLsoil that was

higher (4.0 mg g-1) though. This increase can be explained by limited analyte ex-

traction efficiencies of 62 ± 10% (n=4), which were somewhat lower than those of

Schierz et al. [102] (approx. 75%). Differences may be explained by differences

in sonication (bath vs. horn) and extraction procedure (single vs. sequential).

The nominal MDLsoil is in the range of traditional MWCNT/BC-determination

methods such as CTO-375 (0.23 mg g-1) [93] or more recently developed meth-

ods such as TGA-MS (0.1 mg g-1) [96]. However all of these are still far above

any currently predicted environmental concentration (e.g., 0.01-0.1 µg g-1 for sed-

iments and 0.01 µg g-1 in soils treated with biosolids [25]).

4.3.8 Further optimization of aF4-MALS

Several improvements of the analytical workflow that were beyond the scope of

this conceptual paper may alleviate the current shortcomings of the method and

lead to better recoveries. Analyte losses of 30-40% during extraction are prob-

lematic considering the expected low environmental concentrations. They may be

minimized using, e.g., other surfactants or solvents, and improved extraction pro-

cedures. Additional (selective) analyte loss that occurs during aF4 could be re-

duced by exploring new carrier compositions, separation conditions or membrane

materials. For example, identifying membranes that are unfavorable for soot but

Chapter 4

142

well suited for CNTs (e.g. the polyether sulfone type shown in Table S4.3), may

be an elegant way to increase aF4 recoveries and gain CNT selectivity.

The measures suggested above may also lead to lower MDLs. Additionally, ex-

tracted analyte amounts may be increased through introduction of selective en-

richment steps (e.g, density gradient ultracentrifugation [347]). Analyte amounts

at the detector may be increased through the use of preparative aF4 injection

volumes or use of smart stream splitting [348].

Another shortcoming is related to the aF4-MALS technique itself: as every envi-

ronmental sample - even the one that is analyte-free - will provide a signal (i.e.,

ρ), there will be a need for some kind of a “baseline” measurement of the uncon-

taminated matrix. The best way to deal with this problem may be the establish-

ment of an aF4-MALS database of MWCNT-free soils, spanning over as many

representative soils/soil types as possible.

4.3.9 aF4-MALS within the larger analytical workflow

We envision the presented method to be part of a larger analytical workflow for

detection and quantification of CNTs in natural samples. Already now, the meth-

od presented is considered useful to accompany exposure studies in soils and

sediments, in which CNTs are present in the range of native BC concentrations.

Ideally, the method may be preceded, e.g., by extraction and isolation of BC by

different means (e.g., CTO-375, [94] see also ESI, and other thermo-analytical

methods [349]), and followed by fraction collection, elemental (e.g., by monitoring

embedded trace catalytic metals[350]) and spectroscopic analyses (such as

NIRS [102], Raman [101]). Automated EM may then deliver orthogonal shape

information, provide additional particle number concentrations, and may, coupled

to X-ray analysis, also be useful for BC identification [351].

4.4 Conclusions

We used aF4-MALS to differentiate between particles of different shapes

(MWCNTs, soot and native soil particles). Different MWCNT aspect ratios as well

as mixtures of soot and MWCNTs were efficiently discriminated. Fractions of

MWCNTs in MWCNT-soot mixtures calculated based on ρ obtained by aF4-

MALS were in reasonable agreement with results from automated EM analysis.


143

Compared with native soil, addition of MWCNT led to increased ρ values, while

addition of soot decreased them. The current MDLs for MWCNT are within the

range of other BC quantification methods, but far above any currently predicted

environmental concentration. To overcome this limitation, further development of

suitable enrichment techniques will be necessary. While the method is currently

limited by a rather narrow dynamic range, natural soils exhibited ρ-values that

were consistently at the lower end, allowing for specific identification of high as-

pect ratio particles, such as MWCNTs. Electron microscopic analysis could sup-

port aF4-MALS by providing orthogonal confirmation as well as quantification ca-

pabilities, if contamination effects and losses over aF4 can be reduced. Overall,

aF4-MALS in combination with EM and image analysis could be a valuable tech-

nique to be integrated into future analytical workflows to confirm MWCNT con-

tents in soils.





Swiss National Science Foundation for financial support. We also would like to

thank the Swiss Federal Office for Environment for instrument funding and An-

dres Kaech and the Center for Microscopy and Image Analysis of the University

of Zurich for EM Support. D. Xanat Flores-Cervantes and Felix Wettstein are

acknowledged for fruitful CNT/aF4-discussions, and Isabel Hilber for help with the

statistical analysis. We also thank the company Postnova Analytics GmbH for

their technical support.

Chapter 4

144


Figure S4.1: Example SEM images of the used Soot/MW CNTs suspended in 2%SDC/0.05%

NaN3.


145

Figure S4.2: Example Debye fits (5 th degree) at maximum peak height for (A) soot and (B )

MW1 in pure suspensions. Corresponding r g (filled circles) distributions and MALS 92°

signals (solid lines) are shown below the fits.

Figure S4.3: Example fractograms of the two differe nt mixtures of Nanosphere™ size

standards (A and B), as well as the resulting calib ration function used to determine r h (C).

Chapter 4

146

Figure S4.4: (A) Fractograms obtained by aF4-MALS w ith shape factor ρ (colored circles)

for different injected masses of MW1. Solid lines r epresent the MALS 92° signal in the re-

spective color. Vertical line indicates the average transition point between void/steric and

normal mode elution. (B) Average shape factor ρ obtained from 50% of the MALS 92° peak

width (colored circles) and its signal to noise rat io (crosses) in relation to the injected MW1

mass. Bars represent the standard deviation of the ρ-values over the selected retention

time window.

Figure S4.5: (A) Fractograms obtained by aF4-MALS w ith shape factor ρ (lines) of MW1,

soot and different mixtures (a-e) of both correspon ding to Figure 2 in the main manuscript.


147

(B) Normalized MWCNT-frequency in the five analyzed mixtures of MW1 and soot over the

retention time. MWCNT frequencies were calculated i n MatLab, comparing ρ-values of the

different mixtures at the individual time points re lative to the pure MWCNTs.

Figure S4.6: Example SEM image (transmission mode) of MW1 suspended in 2%

SDC/0.05% NaN3 before fractionation. Note that lengths >1µm are p resent.

Chapter 4

148

Figure S4.7: Fractograms obtained by aF4-MALS with shape factor ρ (symbols) for differ-

ent soils and a sediment. Values represent the aver age of three independent measure-

ments. Vertical lines indicate the transition point between void/steric and normal mode

elution of the respective soil (colors).

Figure S4.8: Fractograms obtained by aF4-MALS with shape factor ρ (symbols) of standard

additions of MW1 (A, average values, n=3) and soot (B, n=1) to a Lufa 2.2 soil extract with

the following analyte concentrations: 0 ( ●), 5 (●), 12.5 (▼), 25 (▲), and 50 (■) µg mL -1 (cor-


149

responding to 1.6, 4, 8.4 and 16.4 mg g -1 of soil, respectively). Vertical line indicates th e

transition point between void/steric and normal mod e elution and solid lines show the 92°

MALS signal of the Lufa 2.2 soil extract (black) an d the highest concentration of MW1 (A,

red) and soot (B, red).

Figure S4.9: Fractograms obtained by aF4-MALS with shape factor ρ (symbols) of standard

additions of MW1 (A, average values, n=3) and soot (B, n=1) directly to soil with the follow-

ing analyte concentrations: 0 ( ●), 1.6 (●), 4 (▼), 8.4 (▲), and 16.4 (■) mg g -1. Vertical line

indicates the transition point between void/steric and normal mode elution and solid lines

show the 92° MALS signal of the Lufa 2.2 soil extra ct (black) and the highest concentration

of MW1 (A,red) and soot (B, red).

Chapter 4

150

Figure S4.10: (A) Alternating injections of Soot an d MW1 on the same membrane. Vertical

line indicates the transition point between void an d normal mode elution. (B) Typical frac-

tograms of the blank 2% SDC/0.05% NaN 3 solution used for dispersion of the samples.

Figure S4.11: Relative standard deviations (%RSD) o f n=3 independent repeated ρ-

measurements of standard additions to a Lufa soil e xtract (A) and directly to soil (B) (cor-

A

B


151

responding to article Figure 4 A and 5 A, respectiv ely). Vertical lines indicate the transition

point between void/steric and normal mode elution.

4.6.1 Example for the integration of aF4-MALS into a different analytical work-

flow: CTO-375

For further illustration, we also analyzed extracts obtained from soil treated by

CTO-375 which is often used to isolate BC (Figure S12 A and B; for information

on the procedure also see Sobek and Bucheli1). When the native BC content is

very low (as in the case of the Lufa 2.2 soil) it can be expected that no difference

in ρ between the four MWCNT concentrations applied here is observed after CTO

treatment of the soil, because they will be far above the BC concentration anyway

(plateau effect, see above). This was confirmed in Figure S12A, where unmodi-

fied MW1 was spiked to an extract of CTO-375 treated soil. Observed ρ-values in

the peak center were comparable to pure MWCNTs. However, when MW1 un-

derwent the complete CTO-375 procedure in soil, some difference in ρ between

the concentrations were observed again (Figure S12B). During CTO-375, the

CNTs are subjected to different oxidizing conditions that may influence their

chemical (e.g., surface functionalization) as well as the physical properties (e.g.,

defects, length). Thus, to thoroughly combine aF4-MALS with CTO-375, addi-

tional knowledge on the transformations that CNTs undergo during the CTO pro-

cess and their effects on CNT behavior in aF4 is required and could be the objec-

tive of future research.

Chapter 4

152

Figure S4.12: Fractograms obtained by aF4-MALS with shape factor ρ (symbols) for (A)

standard additions of MW1 to an extract of a CTO tr eated soil and (B) extracts of a MW1-

spiked soil treated with CTO-375 (for details see m aterials and methods section). Values

represent the average of three independent replicat e measurements. Vertical lines signify

the transition point between void/steric and normal mode elution and solid lines the 92°

MALS signal of the soil (black) and the highest MW1 concentration (red).

Figure S4.13: Example SEM images of Lufa 2.2 soil e xtracts before and after CTO. Before

CTO, mostly organic material is visible, but also s ome Al-Silicates, as shown below (Meas-

ured using EDX). After CTO, extracts are mainly cle an.


153

Table S4.1: Properties of the carbon nanotubes used in the experiments.

Table S4.2: Properties of the used soils.

MWCNT Nominal

TOC

(g/kg

dw)

Length

(µm)

OD

(nm)

ID

(nm)

Max.

aspect

ratio

(nominal)

Functional i-

zation

(wt%)

SSA

(m2/g)

BC 375 °C

(g/kg)

MW1 983.5 10-30 20-

30

5-10 1500 Pristine >110 658±28

MW2 974.6 10-30 ≤8 2-5 3750 3.86% COOH >500 567±0.5

MW3 974.6 0.5-2 ≤8 3-5 250 Pristine >500 239±20

MW4 974.9 0.5-2 ≤8 3-5 250 4-5% COOH >500 715±5.3

Soil name Type TOC

[mg/g]

BC

[mg/g]

LUFA 2.2 Loamy sand, agri-

cultural soil

10.0±2a 0.2 (n=1)b

NIST SRM

1941b

Marine sediment 31.8±3.5c 5.1±1.2d

BC Vertisol Clay vertisol 30.6±1.8c 1.0±0c

(0.8±0.5d)

NABO 1 Clayey loam,

grassland soil

38e 2.1±0.2f

NABO 46 Loamy

clay,agricultural

soil

28e 1.3±0.04f

NABO 67 Organic, vegetable

gardening

261e

11.6±1.5f

NABO 89 Organic, Turf 366e 4.2±0.3f

a)as provided by the distributor, b)this work, c)Ref. [94], d)Ref.

[337], e)Ref. [352], f)Ref. [93]

Chapter 4

154

Table S4.3: %Recovery over different membrane types for MW1 and soot. Values for re-

generated cellulose are average of three independen t measurements. Other membrane

materials have been determined only once (last of t hree subsequent injections).

Membrane MW1 Soot

Regenerated Cellulose

(RC) 10kDa (n=3)

50 50

Polyvinylidenfluoride

(PVDF) 30kDa (n=1)

55 45

Polyethersulfone (PES)

10kDA (n=1)

81 32

Cellulose triacetate

(CTA) 10kDa (n=1)

n.d. n.d.

155

An initial approach to obtain Chapter 5

number based concentrations of CNTs

using automated electron microscopy

image analysis

5.1 Motivation

To define a material as a NM, knowledge on the distribution of its dimensions is

needed, e.g., a number based size distribution in the EU definition [8]. Also, in

terms of toxicology, a relation between the size or the number of particles and a

corresponding effect is of high interest. Many NM sizing methods rely on model

assumptions that relate to spherical particles only. This is the case for example in

dynamic light scattering, where the hydrodynamic radius is determined as an

equivalent radius of a sphere with a specific diffusivity. From such measure-

ments, number based distributions can be easily calculated for particles that are

indeed spherical, such as gold or silver NMs. However, for aspherical particles

with heterogeneous length distributions such as CNTs these assumptions fail,

and number based concentrations cannot be generated. Length distributions of

CNTs have been obtained by aF4-MALS [124] and imaging techniques such as

atomic force microscopy (AFM) [353] and EM [354]. In Chapter 4 it was shown

that automated EM can differentiate between MWCNTs and soot particles, as did

aF4-MALS. However, detection limits in aF4-MALS were still rather high com-

pared to predicted environmental concentrations. This might be addressed by EM

that is in essence able to count individual particles in a sample and also provide

accurate measurements of their 2d-dimensions. The sensitivity of this technique

could be increased compared to aF4-MALS, as particles can be concentrated on

a grid surface using centrifugation, with each single particle deposited being

Chapter 5

156

counted. In addition, the possibility to automate both EM image acquisition and

subsequent image analysis combined with an optimized sample preparation

could potentially speed up the analysis significantly, opening the way for routine

procedures in the future.

For this reason, here, a relatively simple approach to determine number based

CNT concentrations is presented. Different known mass-based concentrations of

pure CNT suspensions were prepared, imaged in transmission mode and ana-

lyzed using ImageJ. The obtained particle numbers were cross-validated by

back-calculation to mass concentrations. Moreover, first images of extracts of

spiked soil samples are presented.


5.2.1 Nanoparticles, dispersions and extracts

The MWCNT powder used in chapter 4 of this thesis and denoted as MW1 was

used for the following experiments. All MWCNT dispersions and soil extracts

were prepared as described in chapter 4, including separation, fraction collection

and recovery determination in aF4. In short, powder samples (MWCNT or soil)

were dispersed in 2% SDC/0.05% NaN3, sonicated and centrifuged at 17.500g.

The resulting supernatant was then used as a working suspension and separated

over aF4.

Differently concentrated suspensions of agglomerate-free MWCNTs were pre-

pared by diluting a 19.3 ppm suspension into 2% SDC/0.05% NaN3 solution with

subsequent sonication for 10 min. in an ultrasonic bath (720W, Bandelin, Berlin,

Germany). The following concentrations were prepared: 0, 1.5, 2.5, 5, 10, 15 and

19.3 µg mL-1. For more realistic samples, MWCNT-spiked soil samples (1.6 and

16 mg g-1) from the study in Chapter 4 were extracted as well.

5.2.2 Sample preparation for automated EM

For sample analysis by automated SEM, 375 µL of the fractions of aF4-runs (the

middle 25% of the peak area) of the abovementioned samples were filled into a

conical Eppendorf tube with a formvar/carbon-coated copper TEM grid (01820,

Carbon Type A, 300 mesh, removable formvar, TedPella, Redding, CA) on a

An initial approach to obtain number based concentrations of CNTs using automated electron microscopy image analysis

157

plastic stopper. This setup resulted in a water column of 5 mm height above the

TEM grid. The particles were deposited on the TEM grids by centrifugation (1h at

16’000 g) using a swinging bucket rotor. For the removal of labile organic con-

taminations from the grid which interfere in the image analysis, a Fischione 1020

plasma cleaner (Fischione, Export, PA) with a 20% oxygen argon plasma was

used.

5.2.3 Automated EM analysis

Images were recorded on a SEM (Nova Nano-SEM 230, FEI, USA) with a trans-

mission electron detector. The bright field signal was used for image formation.

Images were automatically recorded in fast scan mode (approx. 5 images per

min) at a fixed magnification (20’000) using INCA suite 4.15 (Oxford Instruments,

Oxford, UK). For each concentration, 50 images were recorded. All images were

automatically processed in Fiji (ImageJ 1.48q).

5.2.4 Image processing

To each raw image, the following steps were applied in an automated fashion in

Fiji: (i) conversion to 8bit grayscale, (ii) subtract background (light background)

and a rolling ball radius of 10, (iii) thresholding at 0.234, (iv) median filter with a

radius of 4 pixels and finally (v) make binary. Particles overlapping with the image

edges were excluded from the analysis.

Figure 5.1: Image processing example at a magnifica tion of 20000x, showing the original

image (A) and the same image after application of a ll processing steps (B).

Chapter 5

158

To obtain length distributions, the thresholded image (as shown in Figure 5.1)

was skeletonized using the ‘Skeletonize 2D/3D’ and ‘Analyze Skeleton 2D/3D’

plugins [338-340] in ImageJ. The “longest-shortest” branch was then used as the

skeleton length.

5.2.5 Calculations

For validation purposes, from the obtained number of particles, the corresponding

mass of the particles was calculated based on the following approach and com-

pared to the initial nominal MWCNT mass in the dispersions:

From the applied image analysis, the area of each particle was obtained. The

geometric shape of the MWCNTs was then simplified into a cylinder with dimen-

sions of:

`a = 2-b ca = d-=b

With Ap being the measured area (2d representation) of the particle in the image,

Vp the particle volume needed for calculation of the mass, r the particle radius

and l the particle length.

Thus we can write

ca = d e`a2b f

=b

As the particle aspect ratio is defined as

`� = bg

With d as the nominal particle diameter, it follows for the particle volume:


159

ca = d `a=(`�g)4

From the MWCNT density it follows for the mass of the particles:

icajklmnopq�jrkl�mn = lajrn

with ρCNT being an average density for CNTs (see below). The concentration of

the particles can then be calculated from the scanned volume, which is:

c��s��t = `u1s.�vu1s.�ℎ�uxyut��y1�

With Aimage being the area of a single image, nimage the total number of images

acquired and hliquid column the height of the liquid above the grid.

The concentration of the particles can then be calculated as follows:

ka = lac��s��t

The following parameters were used for the calculation:

d = 25 nm (average nominal outer diameter, according to the manufacturer)

ρCNT =1.7 g cm-3 (MWCNT density expected at the selected outer diameter, ac-

cording to Laurent et al.[355])

Aimage =48.7 µm2

hliquid column = 5 mm

Chapter 5

160


5.3.1 Potential for the optimization of sample preparation and image processing

The main problems encountered during sample preparation were associated with

contaminations of the grid. Among these were often occurring diffuse carbon

depositions on the sample surface exposed to the electron beam (electron beam

induced deposition (EBID), visible as a darkening “black box” [356]) as well as

special contaminations originating from the surfactant used for MWCNT disper-

sion (drying artefacts) and general sample handling. One way to overcome con-

taminations from surfactant drying artefacts was the use of aF4 with a surfactant

free carrier solution before preparation of the sample. This resulted in usually

very clean samples and images with clearly visible MWCNTs, while preparation

of the initial suspension in 2% SDC showed high organic contaminations that

could not be removed by plasma cleaning (PC). Formation of EBID however

could be suppressed by the use of PC. MWCNTs survived 10-30s of PC. The

limiting factor in the use of PC is the thermal stability of the TEM-grid. Depending

on the film material, the manufacturer and even the lot, thermal stability of the

grids varies, resulting in maximum PC time of 30s (with the grid type described in

the materials and methods). However, after 30s, already some cracks in the

formvar film were visible, thus a PC time of 15s was used in the following. Anoth-

er option that could help to reduce the amount of various organic contaminations

- but was not tested yet - could be to heat the sample more gently under vacuum.

Also, TEM-grids build from more thermally stable materials such as silicon nitride

could enhance possible PC times and thus improve image quality.

Regarding analysis times, the main drawback to this end is the use of aF4 with

run times of 1 h per sample. Sample preparation for EM is usually quite fast; de-

pending on the centrifuge rotor used approx. 30 grids can be prepared together

per centrifuge run (1h). Then, approx. 50 images can be acquired from each

sample in 10 min using fast scan mode.

During image analysis, it was noted that in some cases CNTs were accidentally

segmented (this can also be seen in Figure 5.1. B). This problem arose from the

fact that one single threshold value was applied to all images in the same fash-


161

ion. The value used was gathered from the optimization of a few images. Howev-

er, the obtained value was obviously not optimal for each image. For future au-

tomation, it is therefore advisable that the threshold value can be set automatical-

ly for each image based on a tailored algorithm (if possible). In addition, thresh-

olding of an image also affects the degree of noise. Here, the use of a median

filter (replaces the value of each pixel by the median value of its neighbours)

proved to be helpful to reduce “salt and pepper”-noise (single (or few)-pixels).

However, still some background remained, as can be seen in the quantitative

section below.

5.3.2 Quantitative evaluation

Number concentrations obtained from image analysis of images as shown in Fig-

ure 5.1B were positively correlated with the expected mass based particle con-

centrations after aF4 (R2=0.85, p=0.024, Figure 5.2A). Using all particles ob-

tained from image analysis for the calculation of the corresponding mass based

concentration a positive correlation was achieved as well (R2=0.79, p=0.017, Fig-

ure 5.2B). The calculated concentrations were mostly in reasonable agreement

with the expected values, showing recoveries of 83-496%, when all particles

were used for calculation (Figure 5.2C). In absolute terms, the detected particle

masses were in the order of pg per analyzed area. However, recovery values

were far above 100% at the three lower concentrations, probably due to higher

background interference in this concentration range. When particles with an as-

pect ratio below 2 were excluded from the calculation, recoveries dropped to 34-

81%. The extent of this reduction was however much more pronounced at the

three lower concentrations, indicating that this step potentially helped to reduce

background interferences. In addition, the correlation between expected and cal-

culated values increased (R2=0.89, p=0.005). For the performed calculations in

general, it is worthy to note that MWCNT density may vary in dependence of the

outer diameter and the number of walls. According to Laurent et al. [355] the

density of MWCNTs in the present outer diameter range (20-30nm) can vary by

approx. 17% and consequently also influence the outcome of such calculations.

Using a more restrictive cut-off (AR>4) resulted in a near to complete loss of par-

ticles in most of the cases.

Chapter 5

162

Figure 5.2: (A) Particle number concentrations quan tified by image analysis in aF4 frac-

tions of the different employed MWCNT concentration s (from images as in Figure 5.1B), (B)

Expected mass based MWCNT concentrations after aF4 in relation to mass based MWCNT

concentrations obtained from automated EM image ana lysis (calculated as described in

the materials and methods section) and (C) obtained recoveries for the calculated mass

based concentrations.

This is in accordance with obtained skeleton length distributions that show that

the samples were dominated by shorter particles (Figure 5.3). An explanation for

this low content of longer particles compared to previous results (see chapter 4)

could be the extremely low particle recovery over aF4 in this case (11±2%) that

led to an extensive loss of the longer particle fraction. The reasons for this low

recovery (also compared to the one achieved in chapter 4) could however not be

clarified. Partially occurring erroneous segmentation of CNTs during image anal-

ysis (see section before) is also likely to have contributed to the observed distri-

bution. In addition, the used MWCNT powder is likely to contain residual soot and

amorphous carbon that contribute to the observed low aspect ratio fraction.


163

Figure 5.3: Skeleton length distribution obtained f rom a 19.3 mg L -1 MWCNT dispersion

after aF4.

5.3.3 First images from extracts of spiked soils

Figure 5.4 shows first images obtained from aF4 fractions of extracts of spiked

soils at low (Figure 5.4A) and high concentration (Figure 5.4B). The latter con-

tained visibly more MWCNTs. Furthermore, it can be positively noted that also

from soil samples, the MWCNTs can be dispersed well enough to potentially en-

able image analysis of single tubes. For the analysis of soil samples compared to

the pure suspensions, it will be essential to ensure a higher recovery of the long-

er particle fraction to increase contrast. It would be beneficial, if a dispersing

agent could be found that is more easily removed from the sample surface than

SDC (e.g., an organic solvent of sufficiently high purity that can be evaporated)

while still guaranteeing full dispersion of the MWCNTs.

Chapter 5

164

Figure 5.4: Scanning transmission electron microsco py micrographs of aF4 fractions of

extracts of the LUFA 2.1 soil, spiked with (A) 1.6 mg g -1 MWCNTs and (B) 16 mg g -1

MWCNTs, respectively. MWCNTs are highlighted with r ed arrows.


165

5.3.4 Conclusion

In this small example, it could be shown that in principle, automated EM image

analysis has the potential to generate reasonable number based concentrations

of pure MWCNT suspensions. With optimized sample preparation – especially if

the use of aF4 can be avoided - analysis times could be fast enough to be prom-

ising for a potential routine use. Method sensitivity is greatly increased compared

to aF4-MALS, with quantified particle masses (calculated) in the pg range. Imag-

es of soil extracts with a suitable quality for image analysis could be generated.

However, the approach is still hampered by the fact that aF4 with low recovery of

long MWCNTs and long run times has to be employed to achieve images suitable

for the actual automated analysis. In addition, image processing must be opti-

mized to avoid accidental segmentation of CNTs and to decrease noise. Explora-

tion of techniques able to remove labile organic carbon contaminations - such as

heating under vacuum - merits further investigations.

Chapter 5

166

167

Titanium dioxide nanoparti-Chapter 6

cles and carbon nanotubes in a soil

mesocosm: vertical translocation in soil

and plant uptake

Alexander Gogos, Janine Moll, Florian Klingenfuss, Marcel van der Heijden,

Fahmida Irin, Micah J. Green, Renato Zenobi and Thomas D. Bucheli

In preparation for submission to the journal “Environmental Pollution”

Chapter 6

168

Abstract

Agricultural soils represent a potential sink for increasing amounts of different

nanomaterials that nowadays inevitably enter the environment. Knowledge on the

relation between their actual exposure concentrations and biological effects on

crops and symbiotic organisms is therefore of high importance. In this part of a

joint companion study, we describe the vertical translocation as well as plant up-

take of three different titanium dioxide (nano-)particles (TiO2 NPs) and multi-

walled carbon nanotubes (MWCNTs) within a pot experiment with homogenously

spiked natural agricultural soil and two plant species (red clover and wheat). TiO2

NPs exhibited limited mobility from soil to leachates and did not induce significant

titanium uptake into plants, although average concentrations were doubled from 4

to 8 mg kg-1 Ti at the highest exposures. While the mobility of MWCNTs in soil

was limited as well, microwave induced heating suggested MWCNT-plant uptake

independent of the exposure concentration.

Titanium dioxide nanoparticles and carbon nanotubes in a soil mesocosm: vertical translocation in soil and plant uptake

169

6.1 Introduction

It is scientifically ascertained that, due to their increased production and use, na-

nomaterials (NMs) will inevitably enter the environment [289], including soils. The

currently most produced NMs are titanium dioxide nanoparticles (TiO2 NPs) [23].

They are used in diverse applications such as paints, UV-protection, photovolta-

ics and photocatalysis [199], but also as a food additive [59]. Carbon nanotubes

(CNTs) are closing the gap in the last years, with 10-fold increased production

volumes since 2006 [24]. Due to their extraordinary mechanical and electrical

properties, CNTs are mostly used as building blocks in light-weight composite

materials as well as electronics.

These particles can enter soils via different pathways [234, 289]. Application of

biosolids to landfills and irrigation with surface waters is most likely for TiO2 NPs,

while CNTs may enter soils via landfills and atmospheric deposition [144]. These

types of release are unintentional, however, also applications in plant protection

and fertilization have been foreseen [324, 357], which may lead to severely in-

creased fluxes of these NP into soils. Apart from the positive effects and func-

tions that are envisioned for agricultural applications of TiO2 NPs and CNTs [324,

357], such as protection of active ingredients and increased plant growth, respec-

tively, also negative effects on microorganisms and plants have been reported

[246, 251, 325].

The enduring uncertainty regarding the environmental safety of NP highlights the

need for a thorough risk assessment of these materials, which includes the study

of their effects on organisms and the ecosystem as well as their fate. However,

the analysis of NPs such as TiO2 and CNTs in complex systems such as real

soils is challenging in many ways. For both, elemental analysis alone is not suffi-

cient to trace the particles due to high elemental background concentrations of Ti

and carbon.

Therefore, most studies until now used simplified laboratory systems as well as

specifically labeled particles for eased detection to investigate both NP transport

through porous media as well as plant uptake, often without confirmation of actu-

al exposure concentrations. For example, TiO2 NP transport was investigated in

sand columns under well controlled conditions [358, 359]. Fang et al. [242] stud-

Chapter 6

170

ied TiO2 NP transport through soil columns at very high concentrations (40 g kg-

1). However, vertical translocation of both TiO2 NPs and CNTs has neither been

investigated yet on a large scale, i.e., large pot experiments or field studies, nor

in the presence of plants. Plant uptake was shown for TiO2 NP in hydroponic ex-

posure systems at high concentrations [360, 361]. In contrast, in a more realistic

exposure setting using natural soil amended with TiO2 NPs, Du et al. [251] found

no uptake of Ti into wheat. Also, CNTs were shown to be taken up into plants

[159, 237, 362] from hydroponic systems. However, until now, no data is availa-

ble for CNT uptake from natural soils, in which CNT transport and subsequent

availability to plants could be different due to their high interactions with the soil

matrix [243, 363, 364].

Here, we investigated the vertical distribution and leaching behavior of three dif-

ferent TiO2 (nano-)particles (P25, E171 and bulk TiO2) and the vertical distribu-

tion of a multi-walled CNT (MWCNT) within two elaborate pot exposure studies

with red clover (Trifolium pratense) [365] and spring wheat (Triticum spp.) [366] in

natural soil, and quantified their fractions in aboveground parts of the plants. We

used recently developed methods such as microwave induced heating (MIH) [98]

and asymmetric flow field-flow fractionation coupled to multi-angle light scattering

(aF4-MALS) [367] to detect and quantify unlabeled MWCNTs in plant and soil

samples, respectively. We additionally imaged root cross sections of exposed

plants using (scanning) transmission electron microscopy. All data from this study

were gathered to accompany two corresponding effect studies with actual, rather

than nominal exposure concentrations. These studies examined the functionality

of the ecosystem in presence of the NPs with regard to nitrogen fixation by the

red clover-rhizobium symbiosis, as well as root colonization by arbuscular mycor-

rhizal fungi of both red clover [365] and wheat [366].


6.2.1 Chemicals and nanoparticles

Food grade E171 TiO2 particles were obtained from Sachtleben Chemie GmbH

(Duisburg, Germany). All other chemicals and TiO2 nanoparticles were purchased


171

from Sigma-Aldrich (Buchs, Switzerland). Uncoated titanium containing NPs were

selected to represent different primary particle size ranges; average primary par-

ticle sizes were determined by TEM image analysis and were 29±9 (P25, n=92),

92±31 (E171, n=52) and 145±46 nm (Bulk TiO2, n=49), see also Figure S6.1. An-

atase was the dominating crystal structure in all of the used particles. However,

P25 also contains 20% rutile, according to the manufacturer.

Multi-walled carbon nanotubes were purchased from Cheap Tubes Inc. (Brattle-

boro, VT). They were declared to have a length of 10-30 µm, and outer diameter

of 20-30 nm, a purity of >95% and an elemental carbon content of >98%. The

MWCNTs were used as received without further purification. Further characteri-

zation of the MWCNTs used was carried out and described in [94, 367]. All pa-

rameters were confirmed to be within the specified ranges with the exception of

CNT length. The latter could only be determined in suspension, where it may

have been altered due to sonication necessary for dispersing the particles.

6.2.2 Soil

A natural soil was collected from an agricultural field at the facility of Agroscope,

Zurich (N47° 25' 39.564" E8° 31' 20.04"). The soil was classified as brown earth

with a sandy loamy to loamy fine fraction. The top layer (5 cm) of the soil was

removed and approx. 0.9 m3 of the underlying 15 cm topsoil were sampled. The

soil was then sieved <5 mm, homogenized by shoveling it three times from one

soil pile to another, and stored in a dry place until it was used in both clover and

wheat experiments.

6.2.3 Spiking of the soil with NPs

The soil was firstly blended with quartz sand (50 % v/v) to facilitate the recovery

of below-ground plant organs after harvest. The properties of the soil-sand mix-

ture are listed in Table 1. First, 300 g of the sand-soil mixture were each mixed

with (i) 0.03 g (wheat experiment only), 0.3 g (clover experiment only), 3 g and 30

g of TiO2 NPs (both experiments), and (ii) 90 mg and 88 g MWCNT powder (clo-

ver experiment only), each in a 500 mL glass bottle which was rotated in a pow-

der mixer (Turbula® T 2 F, Willy A. Bachofen AG, Basel, Switzerland) for 30 min.

For P25 and MWCNTs, the highest particle amounts resulted in a volume too big

Chapter 6

172

for the glass bottles. Therefore, these were split in two and four aliquots, respec-

tively, and each aliquot mixed with 300 g sand-soil mixture.

Into a cement mixer, 30 kg (including the pre-mixture) of a fresh sand-soil mixture

(50% v/v) were added, to yield final nominal NP concentrations of 1, 10, 100, or

1000 mg kg-1, respectively, for TiO2 NPs, and 3 or 2933 mg kg-1 for MWCNTs.

The mixing chamber was covered with a plastic sheet to avoid dust formation and

run for 6 h. The soil was not dried before mixing to avoid changes to the microbial

community structure, also investigated in Moll et al. [366]. Actual exposure con-

centrations were verified by X-ray fluorescence spectroscopy (XRF, for TiO2) and

chemo-thermal oxidation at 375°C [94] (CTO-375, for MWCNTs/Black Carbon

(BC)) analysis as described below.

6.2.4 General experimental design

A detailed description of the general setup, design and execution of the underly-

ing exposure experiments is given in [365, 366]. In brief, for each plant type sev-

en pot replicates were generated for each NP treatment, consisting of seven

plants per pot for red clover and three for wheat. Each pot was filled with a drain-

age layer of sand (0.5 L, 520 g) and 3.3 kg soil (corresponding to 2.9 L). Each pot

was kept at 50-60% (wheat) and 60-70% (red clover) of the total water holding

capacity (WHC, Table 1) during the entire experiment. Plants were grown over a

period of three months in a greenhouse with a 16 h light period (light intensity of

300 W m-2) and a 25/16°C light/dark temperature regime.

Table 6.1: Properties of the soil-quartz mixture (5 0:50 v/v) administered to the pots.

Parameter Value StDev

Org. C % 0.55 0.03 CEC mmol+/kg 6 CaCO3 % 2.6 pH 7.7 max. WHC g H2O/g dry soil 0.308 Sand % 86.1 0 Silt % 6.3 0 Clay % 6.7 0.5


173

6.2.5 Sampling of soil cores

Soil cores were sampled at the day of harvest from each pot using a conventional

soil driller with a 2 cm diameter. Two cores were taken per pot and each divided

into three depths (0-5, 5-10 and 10-15 cm). For each depth, both subsamples

were joined into one and stored in plastic bags at 4°C until further processing.

6.2.6 Titanium analysis in soils with XRF

The soil samples from the cores were dried at 60°C until a constant weight re-

sulted, and ground to a fine powder using a Retsch ZM400 Ball Mill (Retsch

GmbH, Haan, Germany) with a tungsten carbide bead at a frequency of 25 s-1 for

5 min. Four grams of ground soil were homogenously mixed with 0.9 g of wax

and pressed to a 32 mm tablet at 15 tons. Tablets were analyzed using an ener-

gy-dispersive XRF spectrometer (XEPOS, SPECTRO Analytical Instruments

GmbH, Kleve, Germany). For correction of matrix effects, standard additions of

the respective material to the soil were performed. For quality assurance we also

analyzed a certified lake sediment reference sample (LKSD1, CANMET Mining

and Mineral Sciences Laboratories, Ontario, Canada) with recoveries for Ti of

>95%.

6.2.7 Titanium analysis in leachates with ICP-OES

A week before harvest, each pot was watered with 520 mL tap water, leading to

approx. 110% WHC. Consequently, 45 mL of leachate were collected through a

valve at the bottom of the pots. The leachate was analyzed on the same day

without any further treatment using inductively-coupled plasma optical emission

spectrometry (ICP-OES) (ARCOS, SPECTRO Analytical Instruments GmbH). For

quality control, an external Ti containing standard solution (ICAL, Bernd Kraft

GmbH, Duisburg, Germany) was analyzed. The instrumental limit of quantifica-

tion for Ti was determined at 22 µg L-1.

6.2.8 MWCNT quantification in soil with CTO-375

The CTO-375 procedure used in this study is described in detail in Sobek and

Bucheli [94]. This method quantifies total soil BC, which also encompasses

MWCNT-carbon. We analyzed the soil samples taken from the cores, as well as

Chapter 6

174

the bulk spiked soil before the experiment. For the latter, six random grab sam-

ples of approx. 10 g were taken from the spiked pile.

Briefly, all soil samples were dried at 105°C until weight constancy was achieved,

and ground to a fine powder as described before. Dry and ground soil was then

weighed into Ag-capsules, and subjected to thermal oxidation at 375°C under a

constant air stream for 24 h. Subsequently, CTO treated samples were fumigated

with concentrated HCl for 4 h, washed, dried, placed into Sn capsules and ana-

lyzed using an elemental analyser (Euro EA, Hekatech, Germany). For quality

assurance, we also analyzed a marine sediment (standard reference material

SRM 1941b, NIST, Gaithersburg, US) with a known BC content. The obtained

values were within ±5% of the BC content of previous own measurements as well

as compared to Hammes et al. [337]. Further quality assurance measures will be

covered in the results and discussion section.

6.2.9 MWCNT analysis of soil with aF4-MALS

The method for MWCNT detection using aF4-MALS is described in detail by

Gogos et al. [367]. Briefly, 120 mg of dry and ground soil from the cores were

extracted with 10 mL of a 2% sodium deoxycholate/0.05% sodium azide solution,

sonicated three times for 10 min using a high power sonication bath (720W, Ban-

delin, Switzerland) and centrifuged at 17.500g for 10 min. The supernatant was

then used as a working suspension. This procedure was performed for each rep-

licate of each soil depth. Afterwards, the replicates of each depth were joined to

form a collective sample and analyzed using aF4-MALS, which generates a

shape factor ρ from the radius of gyration and the hydrodynamic radius for each

time point in the aF4 fractogram. The difference in ρ (∆ρ) compared to native soil

is then used to detect the MWCNTs [367]. The method detection limit (MDL) of

the present study is presented and further discussed in the results and discussion

section.

6.2.10 Titanium analysis of plants with ICP-OES

From both plants, the parts used as food or feed were analyzed, i.e. the whole

aboveground clover, and the wheat grains. Dried plant samples were ground to a

fine powder using a Retsch ZM200 centrifugal mill (Retsch GmbH). Subsamples


175

(100 mg) were digested in a mixture of 0.2 mL hydrofluoric acid, 1.5 mL nitric acid

and 0.2 mL hydrogen peroxide using a microwave (Ultraclave, MLS, Germany).

The sample volume was subsequently adjusted to 50 mL. Digested samples

were analyzed using ICP-OES (CIROS, SPECTRO Analytical Instruments

GmbH). For quality assurance we also analyzed an industrial sludge reference

sample (standard reference material SRM 2782, NIST, Gaithersburg, US) with

recoveries for Ti of >85%.

6.2.11 MWCNT analysis of plants with MIH

Dry plant material was ground to a fine powder as described before. The amount

of MWCNT uptake was then quantified by MIH, which is described in detail by Irin

et al. [98]. (MW)CNTs have a high microwave absorption capacity, which results

in a rapid rise in temperature within a very short microwave exposure time. Origi-

nal method development included the generation of a calibration curve using the

thermal response as a function of known CNTs spiked into Alfalfa (Medicago sa-

tiva) root samples.

Utilizing the data from Irin et al. [98], a new calibration curve was generated,

where the slope of the curve depends on the respective nanomaterial and the

intercept on the sample type. To this end, first, the initial slope was corrected us-

ing a factor based on the ratio of the source nanomaterials (MWCNTs of this

study) microwave sensitivity and the one of the Irin et al. study. The sensitivity

was determined by exposing ~1 mg of MWCNT powder to 30 W microwave pow-

er (2.45 GHz frequency) and recording the final temperature rise immediately

(within 1 s) with a temperature rise (∆T) of 346 °C. Second, the intercept was cor-

rected based on the control plant microwave response. Figure S6.2 shows the

renormalized calibration curve for MWCNTs at 50 W (6 s). The plant samples

from the controls and the two MWCNT treatments were then tested at 50 W over

6 s and the quantity of MWCNT uptake were calculated using this new calibration

curve. The limit of detection (LOD) as well as the limit of quantification (LOQ)

where calculated based on the temperature rise from five measurements of con-

trol plant samples (blank signal) according to Keith et al. [342] (3 and 10σ above

the blank signal, respectively).

Chapter 6

176

6.2.12 Transmission electron microscopy of root cross sections

Fresh root samples were washed with tap water and pre-fixed in 2.5% glutaral-

dehyde in phosphate buffered saline directly on the day of harvest and stored at

4°C until processing. Ultrathin cross sections (70 nm thickness) were obtained by

cutting root samples embedded in epoxy-resin using an ultramicrotome (Ultracut

E, Leica, Wetzlar, Germany). The detailed sample preparation steps are provided

in the SI. Ultrathin sections were imaged using a TEM (Tecnai G2 Spirit, FEI,

Hillsboro, USA), coupled to an energy-dispersive X-ray (EDX) spectroscope (X-

Max, 80mm2, Oxford Instruments, Abingdon, UK) as well as a STEM (HD-2700-

Cs, Hitachi, Japan) coupled to an EDX system as well (EDAX, NJ) .

6.2.13 Statistics

In the case of normal distributed residuals and homogenous data, an analysis of

variance (ANOVA) was applied. If these model assumptions were not fulfilled, a

Mann-Whitney test was conducted. All statistical analyses were done with the

software R (version 3.01, the R Foundation for Statistical Computing) integrated

in RStudio (version 0.97.551, RStudio, Boston, MA).


6.3.1 Vertical soil distribution and leaching of Ti

Only the highest exposure concentration (1000 mg kg-1) was analytically accessi-

ble using XRF, i.e., standard deviations among the replicates were in the order of

the added Ti amount in samples spiked with <1000 mg kg-1 TiO2. Actual dry

weight exposure concentrations of Ti were almost always slightly higher at the

time of harvest than the initial nominal ones predicted from native and added Ti

amounts, probably due to the residual water content in soils at the time of spiking

(Figure 6.1B/C/E/F). However, the differences were minimal (2.5-7.6%) and

overall not statistically significant (except for Figure 6.1C, P25 1000 mg kg-1, 5-10

cm), indicating that the employed spiking procedure was rather reliable. The con-

trol soils in the wheat experiment were systematically - though not significantly -

lower in Ti content and showed higher standard deviations compared to the con-


177

trols in the clover experiment. This unexpected result may be explained by the

fact that the two experiments were conducted independently using different sub-

sets of the native soil and also highlights the necessity to verify actual exposure

concentrations.

Figure 6.1: Vertical distributions of elemental Ti as determined by XRF analysis for three

depths and for two different exposure experiments: (A-C) Clover controls and clover ex-

posed to 1000 mg kg -1 of Bulk TiO 2 and P25 and (D-F) Wheat controls and wheat exposed

to 1000 mg kg -1 of E171 and P25. Error bars show the standard devi ation of seven repli-

cates. Red squares show the predicted concentration s based on the control values and the

nominal amount of Ti that was added as TiO 2 NPs.

Chapter 6

178

No statistically significant difference could be found between the different soil

layers in any of the treatments (Figure 6.1). Still, some trends could be observed;

the distribution profiles of Ti in the control and in the P25 (80% anatase, 20% ru-

tile) treatments were similar, with a tendency to slightly higher concentrations in

the middle layer in both clover and wheat pots. In contrast, the distribution pro-

files of the two pure anatase particles (Bulk and E171) both tended towards ele-

vated concentrations in the lowest part.

In addition, Ti concentrations in leachates of these two treatments were signifi-

cantly elevated compared to the controls (Figure 6.2, p<0.05), thus it can be as-

sumed that the elevated Ti originated from eluting TiO2 NPs. However, the

leached Ti amount - even in the treatments showing significantly higher concen-

trations – was very low and constituted not more than 10-4% of the initial spiked Ti

amount. In a spiked soil with similar properties to the one used in this study (see

Fang et al. [242], denoted as “JS soil”), a breakthrough of Ti started to occur after

1 pore volume. In our case, 0.52 L of water was added to the pots (equivalent to

30 mm of precipitation) to collect the leachate, which correspond to 0.4 pore vol-

umes only (1.24 L pore volume at full WHC). Thus, the added water amount was

too low to initiate quantitative elution and would therefore explain the relatively

low Ti concentration in the leachate after collection.


179

Figure 6.2: Boxplots (solid line=median) showing th e Ti content of the leachates in the clo-

ver (A, each treatment n=7) and wheat (B, each trea tment n=6) experiment. The LOQ is in-

dicated with a solid red line. Significant differen ce (p<0.05) of a treatment compared to the

respective controls is indicated with an asterisk. The lower and upper borders of the boxes

represent the 25 th and 75 th percentile, respectively. Whiskers represent maxim um and min-

imum values, cirles indicate outliers.

The observed difference in mobility (both in terms of Ti profiles and leachate con-

tent) may partly be explained by differences in the isoelectric point (IEP) of the

TiO2 particles: while the more mobile Bulk TiO2 and E171 exhibited a very low

IEP of 2.2 (see Figure S6.3), the one of P25 was 5.1, being much closer to the

soil pH (7.7, see Table 1) and indicating a lesser colloidal stability [368]. TiO2 NPs

with low IEPs may thus have a higher tendency to reach the groundwater and

should thus be avoided in applications where this might be of relevance, e.g.,

when used as a component of a plant protection product [324, 357].

6.3.2 Vertical soil distribution of BC/MWCNTs

Figure 6.3 shows the BC distribution as well as the shape factor difference (∆ρ)

for the different soil depths of the 2933 mg kg-1 MWCNT amended clover pots. As

Chapter 6

180

with Ti, only the highest MWCNT concentration was analytically accessible. The

total background BC in the control soil was 0.50 ± 0.06 mg g-1 (n=4). The specific

recovery of the employed MWCNT in the soil over the CTO-375 method was

85±13% (n=18, determined by standard addition). Therefore, the expected total

BC concentration in the 2933 mg kg-1 MWCNT amended pots after CTO-375 can

be calculated as follows: (2933x0.85)+500=2993 mg kg-1. However, the average

BC content in the spiked soil before filling into the pots was lower than expected,

with 2400±100 mg kg-1 (n=6), corresponding to 80% of the expected BC concen-

tration. Eventually, losses during the large scale mixing procedure could have

contributed to these lower values. The variability of 4% however suggests that

the employed spiking procedure still resulted in a rather homogenous MWCNT

distribution before the experiment. After the experiment, the average BC content

quantified over all soil depths was 2330±280 mg kg-1, corresponding to 78±12%

(n=15) of the total expected BC concentration, with no significant difference be-

tween the layers. The average value was comparable to the BC content quanti-

fied before the experiment. However, precision, expressed by relative standard

deviations, increased from 4% (original spiked soil) to 12% (aged soil). This in-

crease in variability of the BC content may be associated with partial transport

and/or aging (i.e. physiochemical modification of the particles, influencing their

survival in CTO-375) of MWCNTs during the experiment.


181

Figure 6.3: Vertical distribution of BC content in the clover pots, determined by CTO-375

(black circles), and of ∆ρ values, indicative of the presence of MWCNT, deter mined by aF4-

MALS (red triangles). The dashed black line shows t he native BC content of the soil, while

the dashed red line shows the MWCNT-free soil basel ine in aF4-MALS ( ∆ρ=0). Error bars

show the standard deviation of five replicates. ∆ρ values were determined once from

pooled extracts of the five replicates.

To orthogonally observe the MWCNT behavior between the different layers with a

second method, we also measured the cores with aF4-MALS [367]. With the soil

of the present study, the MDL was at a ∆ρ of 0.099, corresponding to a CTO-

determined MWCNT content of approx. 2 mg g-1 (Figure 6.3), which is slightly

lower than with the soil used in Gogos et al. (4 mg g-1) [367]. The soil layers

showed ∆ρ values of 0.078, 0.1409 and 0.0939 in descending order (Figure 6.3).

Thus, only the value of the middle layer was above the MDL. In combination with

the results from CTO-375 and the increase in variability compared to the initial

spike, this suggests a limited transport of the MWCNTs in the experiment. Such a

Chapter 6

182

low mobility would be in accordance to a dedicated soil transport study by Kasel

et al. [363].

6.3.3 Plant uptake of Ti

With 4.1 mg kg-1, the determined Ti concentration in the clover control plant mate-

rial (Figure 6.4A) was in the range of literature values for a plant species of the

same family (Medicago sativa, a legume which also forms a symbiosis with rhi-

zobia) and total soil Ti [369]. After treatment with TiO2 NPs, the average shoot Ti

content of the clover plants increased to 8 mg kg-1 at the highest exposure con-

centration of both Bulk TiO2 and P25 (Figure 6.4A), however not on a statistically

significant level. For Bulk TiO2, the average Ti content was rising with the expo-

sure concentration, whereas for P25 no such trend could be observed.

Figure 6.4: Total Ti concentration in clover shoots (A) and wheat grains (B) for the different

soil exposures. Error bars indicate one standard de viation (n=4). Different letters above the

bars indicate significant statistical difference (p <0.05).

To elucidate whether this increase in Ti content within the clover shoots was re-

lated to the uptake of actual TiO2 NPs, we investigated cross sections of these

roots with TEM and EDX elemental analysis. In clover roots treated with Bulk

TiO2, Ti containing particles with a similar morphology to the employed particles

(Figure S6.1A) were observed at the root surface (Figure 6.5A/A1) but never in-

side the root cells. Some of these particles also contained Si (Figure 6.5A1, Par-

ticle 2) pointing to a possible natural origin of the particles. However, the absence

of Bulk TiO2 particles within the investigated thin sections does not necessarily


183

disprove particle uptake, as it is not possible to representatively sample a whole

plant root in this way.

In clover roots treated with P25, only very few Ti containing nano-sized particles

were found inside plant cells. The particle B1 in Figure 6.5 shows a clear Ti EDX

peak and is morphologically similar to the employed P25 particles (elongated

hexagon/Figure S6.1C). In addition, the oxygen peak in particle B1 is more dis-

tinct than in the other particles/objects, suggesting that the particle may consist of

titanium-oxide/dioxide.

With an average of 3.3 mg kg-1, the Ti content in the control wheat grains was

slightly lower compared to clover. In this case however, after treatment with TiO2

NPs, the average Ti content in the grains remained approx. constant (Figure

6.4B). Thus, both for clover shoots and wheat grains, no significant difference in

Ti uptake between the different treatments and the controls could be found.

Chapter 6

184

Figure 6.5: Electron microscopy micrographs of (A) an ultrathin-section of a root treated

with 1000 mg kg -1 Bulk TiO 2 (imaged with TEM) together with a magnification (A 1, outside

of the root) and corresponding EDX spectra of selec ted spots (Spectrum 1 and 2) and (B)

an ultrathin-section of a root treated with 1000 mg kg -1 P25 TiO2 (imaged with STEM) to-

gether with a magnification (B1, inside of the root ) and corresponding EDX spectrum of the

selected particle. C represents a particle at a loc ation different from B, but also inside a

cell. EDX spectra were collected from the center of the particles. The copper (Cu) peak that

is present in all spectra originates from the grid material.

While no data is available for clover plants, Larue et al. [370] and Servin et al.

[360] demonstrated that nano-TiO2 can be taken up into wheat and cucumber,

respectively, under extreme conditions (direct hydroponic exposure, high concen-

trations). Larue et al. [370] reported contents of up to 109 mg kg-1 Ti inside wheat

roots, whereas Ti content in wheat leaves was below their LOD. To date, quanti-

tative uptake data for aboveground plant material grown in natural TiO2 NP

spiked soil however is available only from one study performed with wheat plants

[251]. Therein, the Ti content of wheat grains was in the same range as in our


185

study, with no significant uptake, confirming our observations. However, only one

exposure concentration was employed (approx. 100 mg kg-1 TiO2 NPs), so no

comparison can be made with regard to concentration dependent trends.

Altogether, our results suggest that Ti (-NP) uptake to clover plants might also be

possible in real soils, however to a much lesser extent compared to hydroponic

systems. The biological data [365, 366], may represent another indirect piece of

evidence, as for all endpoints (root and shoot biomass, number of flowers, nitro-

gen fixation and arbuscular mycorrhizal colonization), no significant effect of the

treatments were observed for both plants.

6.3.4 Plant uptake of MWCNTs

Figure 6.6 shows the temperature rise (∆T, °C) of dry clover shoot material from

the two MWCNT treatments. The LOD of the MIH method [98] was calculated to

be at ∆T=76°C (corresponding to a 16 µg g-1 MWCNT content) and the LOQ at

∆T=117°C (corresponding to a 55 µg g-1 MWCNT content).

A large fraction of the values was located in the region between LOD and LOQ,

and can thus be considered as MWCNT detections (60% of the values in case of

the 3 mg kg-1 treatment and 43% in case of the 2933 mg kg-1 treatment). The val-

ues above the LOQ represent MWCNT contents of 68 (3 mg kg-1 treatment, n=1)

and 99 µg g-1 (2933 mg kg-1 treatment, n=1).

Chapter 6

186

Figure 6.6: Boxplots (mean=dashed, median=solid) of the temperature increase at 50W, 6s

for the clover plant samples of the two MWCNT treat ments (3 and 2933 mg kg -1). The LOD

(at ∆T=76°C, corresponding to 16 µg g -1) is indicated with the red dashed line and the LOQ

(at ∆T=117°C, corresponding to 55 µg g -1) is indicated with the solid red line. Both LOD an d

LOQ have been determined 3 and 10 σ above the blank signal (control plants), respectiv ely.

All seven replicates of the treatments have been me asured at least twice. The total number

of measurements is indicated above the respective b oxplot. The lower and upper borders

of the boxes represent the 25 th and 75 th percentile, respectively. Whiskers represent maxi-

mum and minimum values.

Taking into account the average dry weight of the clover plants (14.3 g for the 3

mg kg-1 treatment and 15.3 g for the 2933 mg kg-1 treatment, see also Moll et al.

[365]), the two cases with values above the LOD would correspond to a total

amount of MWCNTs of 0.97 mg and 1.5 mg taken up into the plants per pot in the

two treatments, respectively. This means that 9.8% of the initial MWCNT amount

in the soil would have been translocated to the shoots in the 3 mg kg-1 treatment.

Conversely, in the 2933 mg kg-1 treatment, only 0.015% of the initial amount

would have been translocated. It is interesting to note that the MWCNT uptake

was independent from the applied MWCNT concentration. In addition, we ob-


187

served that within the MWCNT treatments, a significant reduction of flowering

occurred (see Moll et al. [365]), which was not concentration dependent as well.

Uptake of CNTs into a plant cell is likely to be limited to the fraction dispersed in

water. MWCNTs however are highly hydrophobic and prone to homo- as well as

hetero-agglomeration with soil constituents. This in turn may result in a very small

fraction of MWCNTs that remains well dispersed in the soil pore water. In addi-

tion, the plant surface may act as a filter that becomes clogged over time. How-

ever, further experiments are needed to explain this intriguing result.

We tried to orthogonally confirm the observed MWCNT uptake by using TEM im-

aging on cross sections of the plant roots. Khodakovskaya et al. [159] and Tripa-

thi et al. [237] provided such optical evidence for CNT uptake from hydroponic

solutions. However, in our case, the sole use of TEM was not conclusive. Figure

S6.4A and S6.4A1 show a MWCNT-like particle that was observed within a plant

root cell of the MWCNT treatment. This particle showed structural and dimen-

sional similarity to the native MWCNTs administered to the pots (Figure S6.4B).

Still, this observation remained the only one within a number of cross sections

that were manually inspected.

We then made additional attempts to screen the samples for the presence of

MWCNTs with confocal Raman spectroscopy (Figure S6.5). However, this ap-

proach requires that the sample is free (or almost free) of carbon allotropes (na-

tive carbon or contaminations), such as soot and amorphous carbon. In principle,

Raman spectroscopy has enough sensitivity to detect single MWCNTs, but we

observed that the spectra of MWCNTs and other carbon allotropes as well as cell

wall material (i.e. lignin [371], which is present in clover roots [372]) had a large

overlap which made the screening difficult.

While the exact amount of MWCNTs taken up could not be fully quantified and

optical confirmation is still not entirely affirmed, based on the specificity of the

MIH method, it is still suggested that MWCNTs were taken up and translocated to

the aboveground part of the plant in some cases. Studies that reported plant up-

take or cellular localization of CNTs until now were performed in hydroponic cul-

tures, where the particles were freely available for interactions with the root [147,

Chapter 6

188

159, 237, 373]. Uptake from soil would thus constitute a novelty, however, due to

the lack of an orthogonal confirmation of the observed uptake, this result should

be interpreted with care.

6.4 Conclusions

As part of a combined effect and exposure study, here we placed emphasis on a

rigorous confirmation of actual NP exposure concentrations. To achieve this goal

we applied an array of analytical techniques to the soil and plant samples, of

which some are novel and used for the first time in this kind of effect studies. In

particular, this includes the combination of CTO-375 and aF4-MALS that showed

that MWCNTs exhibited a rather limited mobility in the soil, as well as MIH that

showed a concentration independent uptake of MWCNTs into some plants. In

addition, the battery of analytical techniques confirmed the relatively constant

exposure situation in both TiO2 NP and MWCNT treatments over several months,

with only subtle changes in concentrations, which could however be explained

qualitatively with underlying NP properties and distribution processes.





Swiss National Science Foundation (SNF) for financial support. Andres Kaech,

Ursula Lüthi and the team at the center for microscopy and image analysis

(ZMB), University of Zurich are gratefully acknowledged for TEM support. We

also thank Jacek Szczerbiński for his support in the confocal Raman microscopy,

Franziska Blum for help with the BC analysis as well as Ralf Kaegi and Brian

Sinnet for their help with the plant Ti determination and the possibility to carry out

the XRF analyses in their lab.


189


Figure S6.1: Bright field TEM micrographs of (A) Bu lk TiO 2, (B) E171 TiO 2 and (C) P25 parti-

cles, as used in the experiments together with corr esponding size information determined

by image analysis (longest distance through the par ticle).

Figure S6.2: Calibration curve for the detection of MWCNT in plant samples. The curve is

derived from the original calibration plot describe d in Irin et al.[98] Note that the intercept

of the curve varies with the heating behavior of th e control sample whereas the slope re-

mains constant for a given nanomaterial.

Chapter 6

190

Figure S6.3: Dependence of the ζ-potential [mV] on pH for the used TiO 2 particles. The dis-

persions were titrated in milli-Q water with hydroc hloric acid/sodium hydroxide solution,

respectively. Intercept with the red line (i.e. 0 m V) denotes the isoelectric point of the parti-

cles. Values represent the average of 3 measurement s (n=3).

Figure S6.4: Transmission electron microscopy micro graphs (A and B), showing (A) a po-

tential CNT structure inside a plant cell (epiderma l cell, part magnified in A1) with structur-

al and dimensional similarities to the MWCNTs admin istered to the pots (B).


191

Figure S6.5: Optical images of root cross sections of controls (A and B) and a MWCNT

treated root (C). The regions indicated with a red square have been magnified for each

optical image and been displayed as a Raman intensi ty map (A1, B1 and C1). The right

panel shows the Raman spectrum of the employed MWCN T powder as well as Raman

spectra taken from indicated regions of the corresp onding Raman maps. The red spectrum

in A1 corresponds to the highest intensity pixel, w hile the blue spectrum represents the

average spectrum of the whole area. The spectra fro m B1 and C1 were taken from the indi-

cated pixels.

Chapter 6

192

6.6.1 Detailed sample preparation steps of root cross sections for analysis using

transmission electron microscopy

Prior to the next steps of the fixation/embedding process, samples were rinsed

three times with PBS. Samples were then incubated with 1% osmium tetroxide at

room temperature for 40 min and then rinsed with water three times. Subsequent-

ly, samples were incubated with 1% uranyl acetate dihydrate in water for 1 h and

rinsed with water three times. Dehydration was performed with 50% (15 min),

70% (20 min), 90% (25 min), 100% (5 min) and 100% water free ethanol for 30

min and propylene oxide 100% for 30 min.

Epon stock solution was prepared by mixing 70.89 g epon 812™, 92.35 g durcu-

pan™ ACM and 8.68 g dibutylphthalate. For the working solution, 5.85 g epon

stock solution were added to 5 g epoxy embedding medium hardener DDSA and

310 mg accelerator DMP 30. Samples were then incubated in 50% epon working

solution and 2 times 100% epon working solution for 1h, respectively. The sam-

ples were polymerized at 60°C overnight.

The resulting epon blocks were pre-trimmed with a razor blade and mounted into

and ultramicrotome (Ultracut E, Leica, Wetzlar, Germany). Ultrathin cross-

sections of the roots (70 nm) were then cut using a diamond knife, transferred to

a formvar/carbon coated copper TEM grid and dried at room temperature.

Conclusions and outlook

193

Conclusions and outlook Chapter 7

7.1 Conclusions

The use of NM to improve agricultural practice is a novel idea, developed over

the last decade. Within the framework of the project “NANOMICROPS”, this the-

sis tried to provide an overview of NMs with potential agricultural use, especially

such envisioned for use in plant protection and fertilization, as well as means for

their analysis in different exposure settings as part of a risk assessment.

It was shown that until 2011, while having increased exponentially since the

2000’s, the measurable activity (i.e. scientific publications, patents and products)

in the field of nano-PPP and fertilizers was marginal in terms of absolute numbers

compared to other fields of nanotechnology. This was in contrast to the public

discussion on the topic, which often exaggerated the actual facts and figures -

both on the positive and the negative side. Such reactions are however not sur-

prising, as the expectations towards a novel high technology - such as nanotech-

nology - are often tremendous and may as well be influenced by strategic aims of

the involved parties (mobilization of campaign support, fund raising etc.) [374].

Due to the enormous diversity of NMs that are actively researched for such appli-

cations, it is impossible to generalize possible benefits, as these depend largely

on the material itself, modifications and the environmental context in which the

NM is used. For example, similar up to lower efficiencies compared to conven-

tional products can be expected especially for metal-based NMs used as active

ingredients, if dissolution and transformation processes that already govern the

activity and the fate of their bulk counterparts cannot be mitigated. In this respect,

as an example the work of Milani et al. [233] shall briefly be mentioned, which

showed similar efficiency of both bulk and nano-ZnO fertilizer in soil. However, if

NMs were used as additives, for example to increase the water solubility of an

Chapter 7

194

active ingredient, to provide a slow release functionality or to protect the active

ingredient, higher efficiencies were often achieved (see also Kah et al. [375]).

The largest fraction of NMs used or actively researched belonged to carbon-

based nanostructures, such as lipid or polymer based materials. Some of these

were designed starting from natural materials such as for example chitosan. As

“persistence” of a chemical (or NM) is an important parameter for risk assess-

ment, such “biocompatible” materials could be of great promise. This observed

trend seems to continue since publication of the study in 2012 (chapter 2). Since

then, several further studies have been published, especially using chitosan-

based NM [376-379] but also using other (biodegradable) polymers [380-383] or

charred nano-structured plant material [384]. If such materials can be generated

(or even “grown”) from biological precursors with a minimum need of other inputs,

this could indeed contribute to a more sustainable agriculture, with reduced use

of fossil resources.

However, regardless of a potential biocompatibility, legal regulation will always

require a thorough risk assessment of a PPP or fertilizer product before approval.

The feasibility of such a procedure is - amongst others - dictated by the availabil-

ity of suitable analytical methods that enable NM characterization and quantifica-

tion in the different test systems over time. It was shown in chapter that for simple

aqueous test systems, which constitute an extensive part of the toxicological ar-

mamentarium, HSI-M could be a useful technique to study and (semi-) quantify

cellular uptake dynamics of NM. Information gathered from HSI-M experiments,

such as subcellular NM localization, agglomeration state, residence time in the

cells and quantity expressed as mapped pixels per cell area can be combined

with toxicological assays (as attempted in chapter 3, supporting information) to

simultaneously elucidate the mechanisms behind possible biological effects. The

technique appears to be most effective in simple and controlled experimental sys-

tems, whose spectral background conditions are well known. This is mainly due

to the limited chemical specificity of the technique that is probably restricted to

metallic particles exhibiting localized surface plasmon resonance (LSPR) [385]. In

addition, it was shown that an influence of the matrix on the spectral profiles is

possible (extracellular substances in this case) and therefore must be taken into

account when acquiring the reference SLs. Already now, HSI-M can display its


195

most obvious strengths, which are its low cost and the fast and easy sample

preparation compared to traditional EM together with the possibility to monitor

unlabeled particles. However, currently the information derived from such meas-

urements remains only semi-quantitative, as the measured values have not been

related to particle mass or numbers, yet.

For more complex test systems such as soils, as mentioned before, a frequent

challenge faced with engineered NMs is that they are often similar in their ele-

mental composition and physical appearance compared to the natural back-

ground - which can be of particulate nature as well. For CNTs, this issue was ad-

dressed in chapter 4, where differences in shape between the CNTs and back-

ground particles were made visible using a combination of aF4 and MALS and

which were then utilized for CNT detection. Although the method was capable of

specifically detecting CNTs in the used soils, it suffered from currently rather high

detection limits (1.6-4 mg g-1). Therefore, this approach is not applicable to real

environmental samples at present, which are expected to contain CNTs at the

lower µg kg-1 level [25]. As the main reason for these high detection limits, losses

of the longer CNTs during sample preparation and especially during aF4 separa-

tion were identified, which limit the dynamic range of the method. However, aF4-

MALS can still be useful to accompany particle exposure and transport studies

that employ high concentrations of unlabeled CNTs (for an example, see chap-

ter 6).

Already in chapter 4, automated EM showed its potential to differentiate between

particles of different shape (i.e. soot and CNTs) in a quantitative manner. There-

fore, and to address the problem of high detection limits, the idea of employing

particle shape as a detection parameter was expanded towards automated EM

which is in essence able to count individual particles, thereby potentially providing

better sensitivity. For pure MWCNT suspensions, it could be shown in chapter 5

that by using automated EM indeed reasonable particle number concentrations in

combination with length distributions can be achieved with an increased sensitivi-

ty compared to aF4-MALS. As the technique could apparently be applicable to

MWCNT-containing soil samples as well, the combined information obtained

could be of use to both toxicologists and environmental chemists in the future.

Chapter 7

196

However, before it can be applied, several methodological obstacles have to be

overcome, including (organic) contaminations, loss of longer CNTs from aF4

separation and incorrect recognition or non-recognition by image analysis.

While some progress with regard to analytical method development could be

achieved as outlined before, still not all of these methods were readily applicable

to the exposure study with plants and real soil that was analytically accompanied

in chapter 6. For example, for HSI-M analysis the employed particles were not

suitable. Carbon nanotubes as black particles possess an inherent unspecific

absorption over most of the analyzed spectral range and TiO2 NPs do not show

LSPR, resulting in rather unspecific spectral profiles (as shown in chapter 3). In

combination with the increased sample complexity compared to the study in

chapter 3, this rendered an application of the technique to both plant and soil

samples very difficult. Also, automated EM was in such an initial stage that it was

not reasonably applicable. However, aF4-MALS together with CTO-375 - a meth-

od developed previously by Sobek and Bucheli [94] - was applied to the soil

samples, showing a limited mobility of the MWCNTs over the experimental peri-

od. Besides the application of aF4-MALS and some previously developed meth-

ods, MIH as a novel analytical technique available through collaboration was ap-

plied to plant samples. It could be shown that, according to MIH, some plants

took up MWCNTs and translocated them to the shoots. The observed uptake was

accompanied by a reduction in flowering, as described by Moll et al. [365]. How-

ever, as confocal Raman microscopy proved unsuitable to unambiguously detect

MWCNTs in the plant samples and no other secondary methods were available,

this result could not be orthogonally confirmed and should be interpreted with

care. Using elemental analysis, TiO2 NPs were shown to exhibit a very limited

mobility from soil to leachates and no statistically significant uptake into the

plants. Both the results for MWCNTs and TiO2 NPs point to a potentially safe use

of these particles in PPP or fertilizers with regard to groundwater contamination

or plant uptake within a short timeframe. However, if such particles were applied

to a field repeatedly, the particles would accumulate in or on the soil with un-

known consequences.


197

7.2 Outlook

Although more and more work is currently being published dealing with nano-

developments in agriculture, the actual market situation along with resilient num-

bers on released quantities still remains somehow obscure. In Switzerland for

example, although since 2010, nano-contents in PPP and fertilizers are mandato-

ry to be declared [386], no nano containing PPP or fertilizer has been registered

yet (personal communication with Dr. Katja Knauer, Federal Office for Agricul-

ture, Berne). In contrast, the French government recently issued a registration

deadline for nano-containing products for May 2014, for which the report was re-

leased end of 2014: approx. 10,000 registrations of nano-containing products

were conducted of which 575 were categorized as phytopharmaceutical (i.e. plant

protection) products [387]. Materials listed in the report under this category in-

cluded especially SiO2, clays, TiO2 and various organic substances. The impres-

sion remains that the topic is still actively being researched while public authori-

ties increasingly show measurable efforts to regulate this novel field.

To support regulation and risk assessment with combined uptake and toxicity

data, the next logical development with respect to HSI-M would be to implement

a system for automated image acquisition and analysis into the current setup.

This would then allow the acquisition of statistically strong data sets which are

indispensable for toxicological studies, where the use of HSI-M would be most

valuable. With the enormous variety of NM types in mind, the capability of HSI-M

to differentiate between them under certain conditions could open up possibilities

for the investigation of e.g., mixture toxicity of different NMs, as recommended for

example by Baun et al. [388]. In addition, the fact that the data are still on a semi-

quantitative level does not preclude the use of HSI-M in such studies. However,

the transition from semi- to fully quantitative data should nevertheless be at-

tempted in the future. The fact that the employed washing procedure for the cell

cultures led to samples where NM containing cells were separated from extracel-

lular NMs could open up the possibility to do so using ICP-based techniques for

example. To count HSI-M “signals” per volume, the use of a hemocytometer with

a reduced thickness specially designed for use in dark field microscopes could be

promising (e.g., a Petroff-Hausser chamber with a 1.4 mm thickness [389]).

Chapter 7

198

To lower the detection limits in aF4-MALS to a level suitable for environmental

risk assessment, losses of the analyte during extraction as well as during separa-

tion in aF4 will have to be mitigated. Therefore, especially further research on

possible extraction procedures (including the selection of appropriate sol-

vents/surfactants) as well as membrane materials (e.g., exploration of completely

new materials or functionalization of existing ones) is needed. First steps in the

latter direction have for example recently been taken by Mudalige et al. [390] for

Au NPs. Another –inherent- limitation of aF4-MALS is given by the fact that all

samples including such that are analyte free provide a ρ value. This makes it

necessary to have “baseline” measurements for uncontaminated matrices. As

shown in chapter 4, this problem may be solved by establishing an aF4-MALS

database of uncontaminated soil types. For orthogonal confirmation and unam-

biguous MWCNT identification, aF4-MALS could be coupled into an analytical

chain together with e.g., monitoring of trace catalytic metals or spectroscopic as

well as electron microscopic characterization.

For the latter, other grid materials than the used formvar/carbon in combination

with heating techniques should be explored for the reduction of (organic) contam-

inations in future studies. Also, the automated image analysis procedure should

be improved to reduce noise and artifacts such as accidental segmentation of

CNTs. In chapter 5, only morphological parameters were used for detection and

quantification. While this may be sufficient for pure suspensions, it may not suf-

fice for real world soil samples. Reducing the samples carbon complexity to con-

tain only BC by applying CTO-375 before CNT extraction could be a first step

(CNTs are still extractable using SDC, as shown in chapter 4, supporting infor-

mation). In the following, centrifugation should be still applied to reduce the

amount of higher density particles. Combining automated image acquisition with

simultaneous automated EDX mapping could then allow at least to exclude all

pixels from the image analysis that contain elements other than C, O and Cu (or

the respective grid material). However, this may still lead to erroneous results, if

the CNTs are not well dispersed and partly associated with such type of inorganic

soil particles. A more sophisticated (but not straightforward) way to access the

desired chemical information (i.e. that of the CNTs), could be to analyze the re-

sulting surface with tip-enhanced Raman spectroscopy (TERS). Pure CNTs show


199

pronounced features within the Raman spectrum; the D-band at 1350 cm-1, G-

band at 1580 cm-1 and the G’-band at 2700 cm-1. While these features are rela-

tively distinct and characteristic for pure CNTs, interpretation of the spectra in

complex systems containing natural particles or other natural entities (e.g., plant

cell walls, see also chapter 6, section 6.3.4) might not be straightforward. Espe-

cially, spectral interferences can be expected for other carbon allotropes present

in natural soils (e.g., soot, see for example Sadezky et al. [391]) but also for other

carbon based materials/substances, such as lignin [371]. This issue of course

also persists with other Raman based approaches, such as the miniaturization

suggested in the following.

Apart from aF4-MALS detection and automated EM, shape selective separation

methods such as deterministic lateral displacement [392] (DLD) coupled to novel

surface enhanced Raman sensors [393], both designed for microfluidic devices,

could constitute an interesting alternative path to follow. A DLD setup would theo-

retically allow continuous separation of differently sized and shaped particles (i.e.

well dispersed CNTs from soil particle background) and for Raman detection, due

to the very small volume in the channels of a microfluidic system, only a potential-

ly very tiny quantity of the analyte would be sufficient for detection. Sample rep-

resentativeness could be ensured through continuous analysis of a larger sample

volume in a flow-through approach.

A general limitation for all analytical methods concerning CNTs arises from the

fact that development of certified reference materials for quality control is still at

its infancy. Recently, in a first approach RM 8281 was released by the National

Institute of Standards and Technology (NIST, Gaithersburg, US) which includes

three length resolved populations of SWCNTs and could be of use for quality

control purposes of the respective methods.

While several methods were available for MWCNT detection and quantification in

chapter 6, no method was at hand to distinguish between Ti originating from the

engineered particles and the natural Ti background from the soil. A possible way

to address this issue in the future could be the determination of the crystal phase

which may differ from one to the other. Most Ti in clays for example is present as

Chapter 7

200

TiO2, and of that, most as the anatase crystal form [394]. The perhaps most in-

vestigated TiO2 NP is P25, which consists of 80% anatase and 20% rutile [395].

X-ray absorption (XAS) Ti K-edge spectra of both anatase and rutile have been

reported to be significantly different from each other [396]. The difference be-

tween these two TiO2 reference spectra and a plant background was sufficient to

determine the Ti speciation in cucumber plants [360] that were grown in P25

amended hydroponic solutions using micro-X-ray absorption near edge spectros-

copy, confirming NP uptake of these NPs from hydroponic solutions. Thus, for

certain soils, XAS measurements might be an opportunity to achieve the goal of

engineered Ti-NP identification in soils.

To obtain the missing orthogonal confirmation of the observed MWCNT uptake, a

correlative (confocal Raman-) microscopy approach would be conceivable. Such

an approach has for example been described for a correlative combination of flu-

orescence light microscopy and cryo-electron tomography [397] as well as for

confocal Raman microscopy and SEM [398]. Provided the availability of a sample

support suitable for both the needs of confocal Raman microscopy and EM, it

could bridge between the interpretational limits of both techniques (see chapter

6), and ultimately provide the necessary unambiguous identification.


201

References

1. ISO/TS 80004-1:2010

2. Nanotechnology definition by the National Nanotechnology Initiative. [cited 2014 14.11.];

Available from: http://www.nano.gov/nanotech-101/what/definition.

3. Nanotechnology definition in Japans Second Science and Technology Basic Plan (2001-

2005) [cited 2014 14.11.]; Available from: http://www8.cao.go.jp/cstp/english/basic/2nd-

BasicPlan_01-05.html.

4. Schummer, J., On the novelty of nanotechnology: A philosophical essay, in In Pursuit of

Nanoethics2014, Springer. p. 15-29.

5. Feynman, R.P., There's plenty of room at the bottom. Engineering and science, 1960.

23(5): p. 22-36.

6. Binnig, G. and H. Rohrer, Scanning tunneling microscopy. Surface Science, 1983. 126(1–

3): p. 236-244.

7. Roco, M.C., C.A. Mirkin, and M.C. Hersam, Nanotechnology research directions for

societal needs in 2020. J Nanoparticle Res, 2011. 13(3): p. 897-919.

8. European Commission Recommendation on the definition of nanomaterial. 2012 [cited

2012 28.02.12]; Available from: http://osha.europa.eu/en/news/eu-european-commission-

recommendation-on-the-definition-of-nanomaterial.

9. Auffan, M., et al., Towards a definition of inorganic nanoparticles from an environmental,

health and safety perspective. Nature Nanotechnology, 2009. 4(10): p. 634-641.

10. Kreyling, W.G., M. Semmler-Behnke, and Q. Chaudhry, A complementary definition of

nanomaterial. Nano Today, 2010. 5(3): p. 165-168.

11. Lövestam, G., et al., Considerations on a definition of nanomaterial for regulatory

purposes2010: Publications Office.

12. Maynard, A.D., Don't define nanomaterials. Nature, 2011. 475(7354): p. 31-31.

13. Qafoku, N.P. and L.S. Donald, Terrestrial Nanoparticles and Their Controls on Soil-/Geo-

Processes and Reactions, in Advances in Agronomy2010, Academic Press. p. 33-91.

14. Buseck, P.R., S.J. Tsipursky, and R. Hettich, Fullerenes from the geological environment.

Science, 1992. 257(5067): p. 215-217.

15. Becker, L., R.J. Poreda, and T.E. Bunch, Fullerenes: An extraterrestrial carbon carrier

phase for noble gases. Proceedings of the National Academy of Sciences, 2000. 97(7): p.

2979-2983.

Curriculum Vitae

202

16. Nowack, B. and T.D. Bucheli, Occurrence, behavior and effects of nanoparticles in the

environment. Environmental Pollution, 2007. 150(1): p. 5-22.

17. Pérez-Arantegui, J., et al., Luster Pottery from the Thirteenth Century to the Sixteenth

Century: A Nanostructured Thin Metallic Film. Journal of the American Ceramic Society,

2001. 84(2): p. 442-46.

18. Nowack, B., H.F. Krug, and M. Height, 120 Years of Nanosilver History: Implications for

Policy Makers. Environmental Science & Technology, 2010. 45(4): p. 1177-1183.

19. Hochella, M.F., M. Spencer, and K.L. Jones, Nanotechnology: Nature's Gift or Scientists'

Brainchild? Environmental Science: Nano, 2015. 2(2): p. 114-119.

20. Gleiter, H., Nanostructured materials: basic concepts and microstructure. Acta Materialia,

2000. 48(1): p. 1-29.

21. Njuguna, J., et al., 1 - Nanomaterials, nanofillers, and nanocomposites: types and

properties, in Health and Environmental Safety of Nanomaterials, J. Njuguna, K.

Pielichowski, and H. Zhu, Editors. 2014, Woodhead Publishing. p. 3-27.

22. http://www.nanotechproject.org/. 2014 [cited 12.01.2015.

23. Piccinno, F., et al., Industrial production quantities and uses of ten engineered

nanomaterials in Europe and the world. Journal Of Nanoparticle Research, 2012. 14(9):

p. 1-11.

24. De Volder, M.F.L., et al., Carbon Nanotubes: Present and Future Commercial

Applications. Science, 2013. 339(6119): p. 535-539.

25. Gottschalk, F., T. Sun, and B. Nowack, Environmental concentrations of engineered

nanomaterials: Review of modeling and analytical studies. Environmental Pollution, 2013.

181(0): p. 287-300.

26. Gottschalk, F. and B. Nowack, The release of engineered nanomaterials to the

environment. Journal Of Environmental Monitoring, 2011. 13(5): p. 1145-1155.

27. Kaegi, R., et al., Synthetic TiO2 nanoparticle emission from exterior facades into the

aquatic environment. Environmental Pollution, 2008. 156(2): p. 233-239.

28. Lorenz, C., et al., Characterization of silver release from commercially available functional

(nano)textiles. Chemosphere, 2012. 89(7): p. 817-824.

29. Farkas, J., et al., Characterization of the effluent from a nanosilver producing washing

machine. Environment International, 2011. 37(6): p. 1057-1062.

30. Benn, T.M. and P. Westerhoff, Nanoparticle Silver Released into Water from

Commercially Available Sock Fabrics. Environmental Science & Technology, 2008.

42(11): p. 4133-4139.

31. Keller, A.A., et al., Release of engineered nanomaterials from personal care products

throughout their life cycle. Journal Of Nanoparticle Research, 2014. 16(7): p. 1-10.

32. Zhang, W.-x. and D.W. Elliott, Applications of iron nanoparticles for groundwater

remediation. Remediation Journal, 2006. 16(2): p. 7-21.

33. Zhang, D., et al., Carbon-stabilized iron nanoparticles for environmental remediation.

Nanoscale, 2010. 2(6): p. 917-919.

References

203

34. Grieger, K.D., et al., Environmental benefits and risks of zero-valent iron nanoparticles

(nZVI) for in situ remediation: Risk mitigation or trade-off? Journal of Contaminant

Hydrology, 2010. 118(3–4): p. 165-183.

35. Tungittiplakorn, W., et al., Engineered Polymeric Nanoparticles for Soil Remediation.

Environmental Science & Technology, 2004. 38(5): p. 1605-1610.

36. Srilatha, B., Nanotechnology in agriculture. Journal of Nanomedicine and

Nanotechnology, 2011. 2(7): p. 123.

37. Sharon, M., A. Choudhary, and R. Kumar, Nanotechnology in agricultural diseases and

food safety. Journal of Phytology, 2010. 2(4): p. 83-92.

38. Perez-de-Luque, A. and D. Rubiales, Nanotechnology for parasitic plant control. Pest

Manag. Sci., 2009. 65(5): p. 540-545.

39. Garcia, M., T. Forbe, and E. Gonzalez, Potential applications of nanotechnology in the

agro-food sector. Ciencia E Tecnologia De Alimentos, 2010. 30(3): p. 573-581.

40. Ghormade, V., M.V. Deshpande, and K.M. Paknikar, Perspectives for nano-biotechnology

enabled protection and nutrition of plants. Biotechnology Advances, 2011. 29(6): p. 792-

803.

41. Colvin, V.L., The potential environmental impact of engineered nanomaterials. Nature

Biotechnology, 2003. 21(10): p. 1166-1170.

42. Boxall, A.B.A., K. Tiede, and Q. Chaudhry, Engineered nanomaterials in soils and water:

how do they behave and could they pose a risk to human health? Nanomedicine, 2007.

2(6): p. 919-927.

43. Choi, O. and Z. Hu, Size Dependent and Reactive Oxygen Species Related Nanosilver

Toxicity to Nitrifying Bacteria. Environmental Science & Technology, 2008. 42(12): p.

4583-4588.

44. Simon-Deckers, A., et al., Size-, Composition- and Shape-Dependent Toxicological

Impact of Metal Oxide Nanoparticles and Carbon Nanotubes toward Bacteria.


45. López-Moreno, M.L., et al., Evidence of the Differential Biotransformation and

Genotoxicity of ZnO and CeO2 Nanoparticles on Soybean (Glycine max) Plants.


46. Lee, W.-M., J.I. Kwak, and Y.-J. An, Effect of silver nanoparticles in crop plants

Phaseolus radiatus and Sorghum bicolor: Media effect on phytotoxicity. Chemosphere,

2012. 86(5): p. 491-499.

47. Hussain, S.M., et al., In vitro toxicity of nanoparticles in BRL 3A rat liver cells. Toxicology

In Vitro, 2005. 19(7): p. 975-983.

48. Poland, C.A., et al., Carbon nanotubes introduced into the abdominal cavity of mice show

asbestos-like pathogenicity in a pilot study. Nature Nanotechnology, 2008. 3(7): p. 423-

428.

49. AshaRani, P.V., et al., Cytotoxicity and Genotoxicity of Silver Nanoparticles in Human

Cells. Acs Nano, 2008. 3(2): p. 279-290.

Curriculum Vitae

204

50. Limbach, L.K., et al., Oxide Nanoparticle Uptake in Human Lung Fibroblasts: Effects of

Particle Size, Agglomeration, and Diffusion at Low Concentrations. Environmental

Science & Technology, 2005. 39(23): p. 9370-9376.

51. Nel, A., et al., Toxic Potential of Materials at the Nanolevel. Science, 2006. 311(5761): p.

622-627.

52. Finizio, A. and S. Villa, Environmental risk assessment for pesticides: A tool for decision

making. Environmental Impact Assessment Review, 2002. 22(3): p. 235-248.

53. Opportunities and Risk of Nanomaterials - National Research Programme NRP64. [cited

22.01.2014; Available from: http://www.nfp64.ch.

54. Moll, J., Effects of nanoparticles on beneficial soil microbes and crops (tentative title),

2015, PhD Thesis, Utrecht University.

55. Pendias, A.K. and H. Pendias, Trace Elements in Soils and Plants1984, Boca Raton, FL:

CRC Press. 315 pp.

56. Hollemann, A.F. and E. Wiberg, Lehrbuch der anorganischen Chemie. Vol. 101. 1995,

Berlin: Walter de Gruyter.

57. Dambournet, D., I. Belharouak, and K. Amine, Tailored Preparation Methods of TiO2

Anatase, Rutile, Brookite: Mechanism of Formation and Electrochemical Properties.

Chemistry Of Materials, 2009. 22(3): p. 1173-1179.

58. Di Paola, A., M. Bellardita, and L. Palmisano, Brookite, the Least Known TiO2

Photocatalyst. Catalysts, 2013. 3(1): p. 36-73.

59. Weir, A., et al., Titanium Dioxide Nanoparticles in Food and Personal Care Products.


60. Yang, Y., et al., Characterization of Food-Grade Titanium Dioxide: The Presence of

Nanosized Particles. Environmental Science & Technology, 2014. 48(11): p. 6391-6400.

61. Fujishima, A., T.N. Rao, and D.A. Tryk, Titanium dioxide photocatalysis. Journal of

Photochemistry and Photobiology C: Photochemistry Reviews, 2000. 1(1): p. 1-21.

62. Fujishima, A. and X. Zhang, Titanium dioxide photocatalysis: present situation and future

approaches. Comptes Rendus Chimie, 2006. 9(5–6): p. 750-760.

63. Hager, S. and R. Bauer, Heterogeneous photocatalytic oxidation of organics for air

purification by near UV irradiated titanium dioxide. Chemosphere, 1999. 38(7): p. 1549-

1559.

64. Matthews, R.W., Purification of water with near—u.v. illuminated suspensions of titanium

dioxide. Water Research, 1990. 24(5): p. 653-660.

65. Horie, Y., et al., Effects of Light Intensity and Titanium Dioxide Concentration on

Photocatalytic Sterilization Rates of Microbial Cells. Industrial & Engineering Chemistry

Research, 1996. 35(11): p. 3920-3926.

66. Kubota, Y., et al., Photokilling of T-24 human bladder cancer cells with titanium dioxide.

British journal of cancer, 1994. 70(6): p. 1107.

References

205

67. Radushkevich, L.V. and V.M. Lukyanovich, O strukture ugleroda, obrazujucegosja pri

termiceskom razlozenii okisi ugleroda na zeleznom kontakte. Zurn. Fisic. Chim., 1952.

26(1): p. 88–95.

68. Monthioux, M. and V.L. Kuznetsov, Who should be given the credit for the discovery of

carbon nanotubes? Carbon, 2006. 44(9): p. 1621-1623.

69. Iijima, S., Helical microtubules of graphitic carbon. Nature, 1991. 354(6348): p. 56-58.

70. Iijima, S. and T. Ichihashi, Single-shell carbon nanotubes of 1-nm diameter. Nature, 1993.

363: p. 603-605.

71. Odom, T.W., J.H. Hafner, and C.M. Lieber, Scanning probe microscopy studies of carbon

nanotubes, in Carbon Nanotubes 2001, Springer. p. 173-211.

72. Ajayan, P.M. and J.M. Tour, Materials Science: Nanotube composites. Nature, 2007.

447(7148): p. 1066-1068.

73. Dresselhaus, M., Carbon connections promise nanoelectronics. Physics World, 1996. 9:

p. 18-19.

74. Wei, B., R. Vajtai, and P. Ajayan, Reliability and current carrying capacity of carbon

nanotubes. Applied Physics Letters, 2001. 79(8): p. 1172-1174.

75. Tans, S.J., A.R. Verschueren, and C. Dekker, Room-temperature transistor based on a

single carbon nanotube. Nature, 1998. 393(6680): p. 49-52.

76. Sun, D.M., et al., A Review of Carbon Nanotube‐and Graphene‐Based Flexible Thin‐Film

Transistors. Small, 2013. 9(8): p. 1188-1205.

77. Dai, L., et al., Carbon nanomaterials for advanced energy conversion and storage. Small,

2012. 8(8): p. 1130-1166.

78. Shannon, M.A., et al., Science and technology for water purification in the coming

decades. Nature, 2008. 452(7185): p. 301-310.

79. Peng, X., et al., Adsorption of 1,2-dichlorobenzene from water to carbon nanotubes.

Chemical Physics Letters, 2003. 376(1–2): p. 154-158.

80. Schwab, F., et al., Diuron Sorbed to Carbon Nanotubes Exhibits Enhanced Toxicity to

Chlorella vulgaris. Environmental Science & Technology, 2012. 47(13): p. 7012-7019.

81. von der Kammer, F., et al., Analysis of engineered nanomaterials in complex matrices

(environment and biota): General considerations and conceptual case studies.

Environmental Toxicology And Chemistry, 2012. 31(1): p. 32-49.

82. Chithrani, B.D., A.A. Ghazani, and W.C. Chan, Determining the size and shape

dependence of gold nanoparticle uptake into mammalian cells. Nano Letters, 2006. 6(4):

p. 662-668.

83. Beach, J., Hyperspectral Imaging. BioOptics World, 2009: p. 2-4.

84. Kent, R.D. and P.J. Vikesland, Controlled evaluation of silver nanoparticle dissolution

using atomic force microscopy. Environmental Science & Technology, 2012. 46(13): p.

6977-6984.

85. Lowry, G.V., et al., Transformations of nanomaterials in the environment. Environmental


Curriculum Vitae

206

86. Kiser, M., et al., Titanium nanomaterial removal and release from wastewater treatment

plants. Environmental Science & Technology, 2009. 43(17): p. 6757-6763.

87. Gondikas, A.P., et al., Release of TiO2 Nanoparticles from Sunscreens into Surface

Waters: A One-Year Survey at the Old Danube Recreational Lake. Environmental


88. Kim, B., et al., Characterization and environmental implications of nano-and larger TiO 2

particles in sewage sludge, and soils amended with sewage sludge. Journal Of

Environmental Monitoring, 2012. 14(4): p. 1128-1136.

89. Kiem, R., H. Knicker, and I. Kögel-Knabner, Refractory organic carbon in particle-size

fractions of arable soils I: distribution of refractory carbon between the size fractions.

Organic Geochemistry, 2002. 33(12): p. 1683-1697.

90. Schnitzer, M. and S.U. Khan, Soil organic matter. Vol. 8. 1975, Amsterdam, The

Netherlands: Elsevier.

91. Harris, D., W.R. Horwáth, and C. van Kessel, Acid fumigation of soils to remove

carbonates prior to total organic carbon or carbon-13 isotopic analysis. Soil Science

Society of America Journal, 2001. 65(6): p. 1853-1856.

92. Gustafsson, O., et al., Evaluation of a protocol for the quantification of black carbon in

sediments. Global Biogeochemical Cycles, 2001. 15(4): p. 881-890.

93. Agarwal, T. and T.D. Bucheli, Adaptation, validation and application of the chemo-thermal

oxidation method to quantify black carbon in soils. Environmental Pollution, 2011. 159(2):

p. 532-538.

94. Sobek, A. and T.D. Bucheli, Testing the resistance of single- and multi-walled carbon

nanotubes to chemothermal oxidation used to isolate soots from environmental samples.

Environmental Pollution, 2009. 157(4): p. 1065-1071.

95. Schmidt, M.W.I. and A.G. Noack, Black carbon in soils and sediments: Analysis,

distribution, implications, and current challenges. Global Biogeochemical Cycles, 2000.

14(3): p. 777-793.

96. Plata, D.L., C.M. Reddy, and P.M. Gschwend, Thermogravimetry-Mass Spectrometry for

Carbon Nanotube Detection in Complex Mixtures. Environmental Science & Technology,

2012. 46(22): p. 12254-12261.

97. Doudrick, K., P. Herckes, and P. Westerhoff, Detection of Carbon Nanotubes in

Environmental Matrices Using Programmed Thermal Analysis. Environmental Science &

Technology, 2012. 46(22): p. 12246-12253.

98. Irin, F., et al., Detection of carbon nanotubes in biological samples through microwave-

induced heating. Carbon, 2012. 50(12): p. 4441-4449.

99. Imholt, T., et al., Nanotubes in microwave fields: light emission, intense heat, outgassing,

and reconstruction. Chemistry Of Materials, 2003. 15(21): p. 3969-3970.

100. Dresselhaus, M.S., et al., Raman spectroscopy of carbon nanotubes. Physics Reports,

2005. 409(2): p. 47-99.

References

207

101. Lopez-Lorente, A.I., B.M. Simonet, and M. Valcarcel, Determination of carboxylic

SWCNTs in river water by microextraction in ionic liquid and determination by Raman

spectroscopy. Talanta, 2012. 105(0): p. 75-79.

102. Schierz, A., et al., Characterization and Quantitative Analysis of Single-Walled Carbon

Nanotubes in the Aquatic Environment Using Near-Infrared Fluorescence Spectroscopy.


103. Badireddy, A.R., M.R. Wiesner, and J. Liu, Detection, Characterization, and Abundance

of Engineered Nanoparticles in Complex Waters by Hyperspectral Imagery with

Enhanced Darkfield Microscopy. Environmental Science & Technology, 2012. 46(18): p.

10081-10088.

104. Brando, V.E. and A.G. Dekker, Satellite hyperspectral remote sensing for estimating

estuarine and coastal water quality. Geoscience and Remote Sensing, IEEE Transactions

on, 2003. 41(6): p. 1378-1387.

105. Manolakis, D. and G. Shaw, Detection algorithms for hyperspectral imaging applications.

Signal Processing Magazine, IEEE, 2002. 19(1): p. 29-43.

106. Baer, D.R., et al., Characterization challenges for nanomaterials. Surface and Interface

Analysis, 2008. 40(3‐4): p. 529-537.

107. Tiede, K., et al., Detection and characterization of engineered nanoparticles in food and

the environment. Food Additives And Contaminants, 2008. 25(7): p. 795-821.

108. Baer, D.R., et al., Challenges in applying surface analysis methods to nanoparticles and

nanostructured materials. Journal of Surface Analysis, 2005. 12(2): p. 101.

109. Koh, A.L., et al., Electron energy-loss spectroscopy (EELS) of surface plasmons in single

silver nanoparticles and dimers: influence of beam damage and mapping of dark modes.

Acs Nano, 2009. 3(10): p. 3015-3022.

110. Giddings, J.C., A New Separation Concept Based on a Coupling of Concentration and

Flow Nonuniformities. Separation Science, 1966. 1(1): p. 123-125.

111. Liu, J., J.D. Andya, and S.J. Shire, A critical review of analytical ultracentrifugation and

field flow fractionation methods for measuring protein aggregation. The AAPS journal,

2006. 8(3): p. E580-E589.

112. Runyon, J.R. and S.K.R. Williams, Composition and molecular weight analysis of styrene-

acrylic copolymers using thermal field-flow fractionation. Journal Of Chromatography A,

2011. 1218(38): p. 6774-6779.

113. Kammer, F.v.d., M. Baborowski, and K. Friese, Field-flow fractionation coupled to multi-

angle laser light scattering detectors: Applicability and analytical benefits for the analysis

of environmental colloids. Analytica Chimica Acta, 2005. 552(1-2): p. 166-174.

114. Dubascoux, S., et al., Optimisation of asymmetrical flow field flow fractionation for

environmental nanoparticles separation. Journal Of Chromatography A, 2008. 1206(2): p.

160-165.

Curriculum Vitae

208

115. Gigault, J., et al., Gold nanorod separation and characterization by asymmetric-flow field

flow fractionation with UV-Vis detection. Analytical And Bioanalytical Chemistry, 2012: p.

1-12.

116. Hagendorfer, H., et al., Characterization of Silver Nanoparticle Products Using

Asymmetric Flow Field Flow Fractionation with a Multidetector Approach - a Comparison

to Transmission Electron Microscopy and Batch Dynamic Light Scattering. Analytical

Chemistry, 2012. 84(6): p. 2678-2685.

117. Tadjiki, S., et al., Detection, separation, and quantification of unlabeled silica

nanoparticles in biological media using sedimentation field-flow fractionation. Journal Of

Nanoparticle Research, 2009. 11(4): p. 981-988.

118. http://www.postnova.com/. [cited 26.11.2014.

119. Ulrich, A., et al., Critical aspects of sample handling for direct nanoparticle analysis and

analytical challenges using asymmetric field flow fractionation in a multi-detector

approach. Journal Of Analytical Atomic Spectrometry, 2012. 27(7): p. 1120-1130.

120. Hagendorfer, H., et al., Application of an asymmetric flow field flow fractionation multi-

detector approach for metallic engineered nanoparticle characterization - Prospects and

limitations demonstrated on Au nanoparticles. Analytica Chimica Acta, 2011. 706(2): p.

367-378.

121. Kammer, F.v.d., et al., Separation and characterization of nanoparticles in complex food

and environmental samples by field-flow fractionation. TrAC Trends in Analytical

Chemistry, 2011. 30(3): p. 425-436.

122. Gigault, J., B. Grassl, and G. Lespes, Size characterization of the associations between

carbon nanotubes and humic acids in aqueous media by asymmetrical flow field-flow

fractionation combined with multi-angle light scattering. Chemosphere, 2012. 86(2): p.

177-182.

123. Claveranne-Lamolère, C., et al., Colloidal transport of uranium in soil: Size fractionation

and characterization by field-flow fractionation-multi-detection. Journal Of

Chromatography A, 2009. 1216(52): p. 9113-9119.

124. Gigault, J., et al., Single walled carbon nanotube length determination by asymmetrical-

flow field-flow fractionation hyphenated to multi-angle laser-light scattering. Journal Of

Chromatography A, 2010. 1217(50): p. 7891-7897.

125. Guan, X., et al., Asymmetric Flow Field-Flow Fractionation with Multiangle Light

Scattering Detection for Characterization of Cellulose Nanocrystals. Biomacromolecules,

2012. 13(9): p. 2671-2679.

126. Mitrano, D.M., et al., Silver nanoparticle characterization using single particle ICP-MS

(SP-ICP-MS) and asymmetrical flow field flow fractionation ICP-MS (AF4-ICP-MS).

Journal Of Analytical Atomic Spectrometry, 2012. 27(7): p. 1131-1142.

127. Podzimek, S., T. Vlcek, and C. Johann, Characterization of branched polymers by size

exclusion chromatography coupled with multiangle light scattering detector. I. Size

References

209

exclusion chromatography elution behavior of branched polymers. Journal of Applied

Polymer Science, 2001. 81(7): p. 1588-1594.

128. Wyatt, P.J., Light scattering and the absolute characterization of macromolecules.

Analytica Chimica Acta, 1993. 272(1): p. 1-40.

129. Wyatt, P.J., Measurement of Special Nanoparticle Structures by Light Scattering.

Analytical Chemistry, 2014. 86(15): p. 7171–7183.

130. Wyatt, P.J., Submicrometer Particle Sizing by Multiangle Light Scattering following

Fractionation. Journal Of Colloid And Interface Science, 1998. 197(1): p. 9-20.

131. Bohren, C.F. and D.R. Huffman, Absorption and scattering of light by small particles

2008, Weinheim, Germany: John Wiley & Sons.

132. Podzimek, S., Light Scattering, Size Exclusion Chromatography and Asymmetric Flow

Field Flow Frationation - Powerful Tools for the Characterization of Polymers, Proteins

and Nanoparticles2011, New Jersey: John Wiley & Sons.

133. Coûteaux, M.-M. and J. F. Darbyshire, Functional diversity amongst soil protozoa.

Applied Soil Ecology, 1998. 10(3): p. 229-237.

134. Som, C., et al., The importance of life cycle concepts for the development of safe

nanoproducts. Toxicology, 2010. 269(2-3): p. 160-169.

135. Knauer, K. and T. Bucheli, Nanomaterials - the need for research in agriculture.

Agrarforschung, 2009. 16(10): p. 390-395.

136. Nair, R., et al., Nanoparticulate material delivery to plants. Plant Science, 2010. 179(3): p.

154-163.

137. Bhattacharyya, A., et al., Nano-particles - A recent approach to insect pest control.

African Journal of Biotechnology, 2010. 9(24): p. 3489-3493.

138. Rashidi, L. and K. Khosravi-Darani, The Applications of Nanotechnology in Food Industry.

Critical Reviews In Food Science And Nutrition, 2011. 51(8): p. 723-730.

139. Barik, T.K., B. Sahu, and V. Swain, Nanosilica - from medicine to pest control.

Parasitology Research, 2008. 103(2): p. 253-258.

140. Wiesner, M.R., et al., Assessing the Risks of Manufactured Nanomaterials.


141. Savolainen, K., et al., Risk assessment of engineered nanomaterials and

nanotechnologies-A review. Toxicology, 2010. 269(2-3): p. 92-104.

142. Ma, X.M., et al., Interactions between engineered nanoparticles (ENPs) and plants:

Phytotoxicity, uptake and accumulation. Science Of The Total Environment, 2010.

408(16): p. 3053-3061.

143. Hendren, C.O., et al., Estimating Production Data for Five Engineered Nanomaterials As

a Basis for Exposure Assessment. Environmental Science & Technology, 2011. 45(7): p.

2562-2569.

144. Gottschalk, F., et al., Possibilities and Limitations of Modeling Environmental Exposure to

Engineered Nanomaterials by Probabilistic Material Flow Analysis. Environmental

Toxicology and Chemistry, 2010. 29(5): p. 1036-1048.

Curriculum Vitae

210

145. Zhu, X.S., Y. Chang, and Y.S. Chen, Toxicity and bioaccumulation of TiO2 nanoparticle

aggregates in Daphnia magna. Chemosphere, 2010. 78(3): p. V-215.

146. Werlin, R., et al., Biomagnification of cadmium selenide quantum dots in a simple

experimental microbial food chain. Nature Nanotechnology, 2011. 6(1): p. 65-71.

147. Lin, S.J., et al., Uptake, Translocation, and Transmission of Carbon Nanomaterials in

Rice Plants. Small, 2009. 5(10): p. 1128-1132.

148. Lopez-Moreno, M.L., et al., X-ray Absorption Spectroscopy (XAS) Corroboration of the

Uptake and Storage of CeO2 Nanoparticles and Assessment of Their Differential Toxicity

in Four Edible Plant Species. Journal Of Agricultural And Food Chemistry, 2010. 58(6): p.

3689-3693.

149. Bourguignon F., et al., Millennium Development Goals at Midpoint: Where do we stand

and where do we need to go?, U. Department for International Development, Editor 2008,

European Commission.

150. Hansen, S.F., A global view of regulations affecting nanomaterials. Wiley Interdisciplinary

Reviews: Nanomedicine and Nanobiotechnology, 2010. 2(5): p. 441-449.

151. Bouwmeester, H., et al., Review of health safety aspects of nanotechnologies in food

production. Regulatory Toxicology and Pharmacology, 2009. 53(1): p. 52-62.

152. Soil density. 04.05.12]; Available from:

http://web.ead.anl.gov/resrad/datacoll/soildens.htm.

153. Mueller, N.C. and B. Nowack, Exposure modeling of engineered nanoparticles in the

environment. Environmental Science & Technology, 2008. 42(12): p. 4447-4453.

154. Ishaque, M., G. Schnabel, and D. Anspaugh, Agrochemical formulations comprising a

pesticide, an organic UV-photoprotective filter and coated metal-oxide nanoparticles,

Patent No. WO/2009/153231, 2009.

155. Guan, H.N., et al., A novel photodegradable insecticide: Preparation, characterization and

properties evaluation of nano-imidacloprid. Pestic. Biochem. Phys., 2008. 92(2): p. 83-91.

156. Gottschalk, F., et al., Modeled environmental concentrations of engineered nanomaterials

(TiO2, ZnO, Ag, CNT, fullerenes) for different regions. Environmental Science &

Technology, 2009. 43(24): p. 9216-9222.

157. Dookhith, M.R. and H. Linares, Stabilisierte Öl-in-Wasseremulsion, Patent No.

DE69126275T2, 1998.

158. Kim, H., et al., Antifungal effectiveness of nanosilver colloid against rose powdery mildew

in greenhouses, in Nanocomposites and Nanoporous Materials Viii, C.K. Rhee, Editor

2008, Trans Tech Publications Ltd: Stafa-Zurich. p. 15-18.

159. Khodakovskaya, M., et al., Carbon nanotubes are able to penetrate plant seed coat and

dramatically affect seed germination and plant growth. ACS Nano, 2009. 3(10): p. 3221-

3227.

160. Miyazaki, K. and N. Islam, Nanotechnology systems of innovation-”An analysis of industry

and academia research activities. Technovation, 2007. 27(11): p. 661-675.

References

211

161. Down on the Farm. 2004 [cited 15.11.2011; Available from:

http://www.etcgroup.org/upload/publication/80/02/etc_dotfarm2004.pdf.

162. Chopra, A., et al. Supplement to the President’s 2012 Budget. 2012 [cited 17.04.2012;

Available from:

http://www.nano.gov/sites/default/files/pub_resource/nni_2012_budget_supplement.pdf.

163. Schrof, W., et al., Nanoparticles comprising a crop protection agent, Patent No.

WO/2003/039249, 2003.

164. Calestous, J. and Y.C. Lee. Innovation: Applying knowledge in development. 2005 [cited

17.04.2012; Available from: http://www.unmillenniumproject.org/documents/Science-

complete.pdf.

165. Jo, Y.K., B.H. Kim, and G. Jung, Antifungal Activity of Silver Ions and Nanoparticles on

Phytopathogenic Fungi. Plant Disease, 2009. 93(10): p. 1037-1043.

166. Min, J.S., et al., Effects of Colloidal Silver Nanoparticles on Sclerotium-Forming

Phytopathogenic Fungi. Plant Pathology Journal, 2009. 25(4): p. 376-380.

167. Park, H., et al., A New Composition of Nanosized Silica-Silver for Control of Various Plant

Diseases. The plant pathology journal, 2006. 22(3): p. 295-302.

168. Alavi, S.V. and D.A. A., Evaluation of the nanosilver colloidal solution in comparison with

the registered fungicide to control greenhouse cucumber downy mildew disease in the

north of iran. Acta Hort., 2010. ISHS(877): p. 1643-1646.

169. Jung, J.-H., et al., The Effect of Nano-Silver Liquid against the White Rot of the Green

Onion Caused by Sclerotium cepivorum. Mycobiology, 2010. 38(1): p. 39-45.

170. Lamsal, K., et al., Inhibition Effects of Silver Nanoparticles against Powdery Mildews on

Cucumber and Pumpkin. Mycobiology, 2011. 39(1): p. 26-32.

171. Rahman Nia, J., Nanosilver for preservation and treatment of diseases in agriculture field,

Patent No. US20090075818, 2009.

172. Goswami, A., et al., Novel applications of solid and liquid formulations of nanoparticles

against insect pests and pathogens. Thin Solid Films, 2010. 519(3): p. 1252-1257.

173. Ulrichs, C., A. Goswami, and M. I., Nano-structured silica—physical active pesticides for

urban settings. Proceedings of the second international symposium on plant protection

and plant health in Europe, 2007.

174. Debnath, N., et al., Entomotoxic effect of silica nanoparticles against Sitophilus

oryzae(L.). Journal of Pest Science, 2010. 84(1): p. 99-105.

175. Stadler, T., M. Buteler, and D.K. Weaver, Novel use of nanostructured alumina as an

insecticide. Pest Management Science, 2010. 66(6): p. 577-579.

176. Schneider, K.-H., et al., Method for treating phytopathogenic microorganisms using

surface-modified nanoparticulate copper salts, Patent No. WO/2011/067186, 2011.

177. Beestman, G.B., Controlled Release in Crop Protection: Past Experience and Future

Opportunities, in Chemistry of Crop Protection2004, Wiley-VCH Verlag GmbH & Co.

KGaA. p. 272-279.

Curriculum Vitae

212

178. Elek, N., et al., Novaluron nanoparticles: Formation and potential use in controlling

agricultural insect pests. Colloids and Surfaces A: Physicochemical and Engineering

Aspects, 2010. 372(1-3): p. 66-72.

179. Anjali, C.H., et al., Neem oil (Azadirachta indica) nanoemulsion—a potent larvicidal agent

against Culex quinquefasciatus. Pest Management Science, 2011: p. n/a-n/a.

180. Da Silva, A., et al., Water Dispersible Agrochemical Formulations, Patent No.

WO/2010/051607, 2010.

181. Giessler-Blank, S., et al., Nanoemulsions and Processes for their Preparation, And Their

Use as Formulations of Plant Protection Agents and/or Pesticides and/or Cosmetic

Preparations, Patent No. US20100041629, 2010.

182. Magdassi, S., B. Dayan, and G. Levi-Ruso, Pesticide Nanoparticles obtained from

Microemulsions and Nanoemulsions, Patent No. WO/2008/032328, 2008.

183. Robinson, D. and G. Salejova-Zadrazilova, Nanotechnologies for nutrient and biocide

delivery in agricultural production, 2010, ObservatoryNANO.

184. Glatter, O., et al., Pesticidal dispersion comprising nanostructured dispersed phase,

Patent No. WO/2011/138701, 2011.

185. Salentinig, S., et al., Transitions in the internal structure of lipid droplets during fat

digestion. Soft Matter, 2011. 7(2): p. 650-661.

186. Stone, V., et al., Nanomaterials for environmental studies: Classification, reference

material issues, and strategies for physico-chemical characterisation. Science Of The

Total Environment, 2010. 408(7): p. 1745-1754.

187. Luo, D.Q., et al., Anti-fungal Efficacy of Polybutylcyanoacrylate Nanoparticles of Allicin

and Comparison With Pure Allicin. Journal of Biomaterials Science-Polymer Edition,

2009. 20(1): p. 21-31.

188. Yang, F.-L., et al., Structural Characterization of Nanoparticles Loaded with Garlic

Essential Oil and Their Insecticidal Activity against Tribolium castaneum (Herbst)

(Coleoptera: Tenebrionidae). Journal Of Agricultural And Food Chemistry, 2009. 57(21):

p. 10156-10162.

189. Botts, M.F., F.C. Kohn, and M.L. Miller, Particles containing agricultural active

ingredients, Patent No. US7070795, 2006.

190. Metselaar, J.M., et al., A Novel Family of l-Amino Acid-Based Biodegradable Polymer-

Lipid Conjugates for the Development of Long-Circulating Liposomes with Effective Drug-

Targeting Capacity. Bioconjugate Chemistry, 2003. 14(6): p. 1156-1164.

191. Vasir, J.K. and V. Labhasetwar, Biodegradable nanoparticles for cytosolic delivery of

therapeutics. Advanced Drug Delivery Reviews, 2007. 59(8): p. 718-728.

192. Yavitz, E.Q., Plant protection and growth stimulation by nanoscalar particle folial delivery,

Patent No. US7494526, 2005.

193. Corradini, E., M.R. de Moura, and L.H.C. Mattoso, A preliminary study of the

incorparation of NPK fertilizer into chitosan nanoparticles. eXPRESS Polymer Letters,

2010. 4(8): p. 509-515.

References

213

194. Li, F.B., H. Pham, and D. Anderson, Methods to produce polymer nanoparticles and

formulations of active ingredients, Patent No. WO/2010/035118, 2010.

195. Majeti, N.V. and K. Ravi, A review of chitin and chitosan applications. Reactive and

Functional Polymers, 2000. 46(1): p. 1-27.

196. Gooday, G.W., Physiology of microbial degradation of chitin and chitosan.

Biodegradation, 1990. 1(2): p. 177-190.

197. Vive la Crop. 2012 [cited 14.05.2012; Available from:

http://www.vivecrop.com/ACCN_April2012.pdf.

198. Syngenta. 2012 [cited 15.5.2012; Available from:

http://www.syngenta.com/global/corporate/en/products-and-innovation/product-

brands/lawn-and-garden/turf-and-landscape/Pages/banner-maxx.aspx.

199. Chen, X. and S.S. Mao, Titanium Dioxide Nanomaterials: Synthesis, Properties,

Modifications, and Applications. Chemical Reviews, 2007. 107(7): p. 2891-2959.

200. Frensch, H., Evaporation inhibiting additive for concentrated dispersions of plant

protection products, 1975, HOECHST AG.

201. Fujishima, A. and K. Honda, Electrochemical Photolysis of Water at a Semiconductor

Electrode. Nature, 1972. 238(5358): p. 37-38.

202. Thomas, J., K. Kumar, and C. Praveen, Synthesis of Ag Doped Nano TiO2 as Efficient

Solar Photocatalyst for the Degradation of Endosulfan. Advanced Science Letters, 2011.

4: p. 108-114.

203. Baruah, S. and J. Dutta, Nanotechnology applications in pollution sensing and

degradation in agriculture: a review. Environmental Chemistry Letters, 2009. 7(3): p. 191-

204.

204. Kumar, S.G. and L.G. Devi, Review on Modified TiO2 Photocatalysis under UV/Visible

Light: Selected Results and Related Mechanisms on Interfacial Charge Carrier Transfer

Dynamics. The Journal of Physical Chemistry A, 2011. 115(46): p. 13211-13241.

205. Kikuchi, Y., et al., Photocatalytic bactericidal effect of TiO2 thin films: dynamic view of the

active oxygen species responsible for the effect. Journal of Photochemistry and

Photobiology A: Chemistry, 1997. 106(1-3): p. 51-56.

206. Lu, J.W., et al., TiO2 photocatalytic antifungal technique for crops diseases control.

Journal of Environmental Sciences-China, 2006. 18(2): p. 397-401.

207. Guan, H.-N., et al., Novel photodegradable insecticide W/TiO2/Avermectin

nanocomposites obtained by polyelectrolytes assembly. Colloids and Surfaces B:

Biointerfaces, 2011. 83(1): p. 148-154.

208. Guan, H., et al., Dynamics of residues from a novel nano-imidacloprid formulation in

soyabean fields. Crop Protection, 2010. 29(9): p. 942-946.

209. Vormberg, R., et al., Preparation containing UV radiation absorbing metal oxide powder

and a superspreading agent, Patent No. WO/2007/014826, 2007.

Curriculum Vitae

214

210. Gao, F.Q., et al., Mechanism of nano-anatase TiO2 on promoting photosynthetic carbon

reaction of spinach - Inducing complex of Rubisco-Rubisco activase. Biological Trace

Element Research, 2006. 111(1-3): p. 239-253.

211. Yang, F., et al., The improvement of spinach growth by nano-anatase TiO2 treatment is

related to nitrogen photoreduction. Biological Trace Element Research, 2007. 119(1): p.

77-88.

212. Linglan, M., et al., Rubisco Activase mRNA Expression in Spinach: Modulation by

Nanoanatase Treatment. Biological Trace Element Research, 2008. 122(2): p. 168-178.

213. Zheng, L., et al., Effect of nano-TiO2 on strength of naturally and growth aged seeds of

spinach. Biological Trace Element Research, 2005. 104(1): p. 83-91.

214. Hong, F.H., et al., Effect of nano-TiO2 on photochemical reaction of chloroplasts of

spinach. Biological Trace Element Research, 2005. 105(1-3): p. 269-279.

215. Wani, I.A., et al., Silver nanoparticles: Large scale solvothermal synthesis and optical

properties. Materials Research Bulletin, 2010. 45(8): p. 1033-1038.

216. El Badawy, A.M., et al., Surface Charge-Dependent Toxicity of Silver Nanoparticles.


217. Baker, C., et al., Synthesis and Antibacterial Properties of Silver Nanoparticles. Journal

Of Nanoscience And Nanotechnology, 2005. 5(2): p. 244-249.

218. Grape diseases - Powdery mildew. 2010, cited 16.01.2012]; Available from:

http://www.agf.gov.bc.ca/cropprot/grapeipm/mildew.htm.

219. Johansson, M., M. Pell, and J. Stenström, Kinetics of Substrate-Induced Respiration

(SIR) and Denitrification: Applications to a Soil Amended with Silver. Ambio, 1998. 27(1):

p. 40-44.

220. Liu, J. and R.H. Hurt, Ion Release Kinetics and Particle Persistence in Aqueous Nano-

Silver Colloids. Environmental Science & Technology, 2010. 44(6): p. 2169-2175.

221. Wang, L., et al., Watching Silica Nanoparticles Glow in the Biological World. Analytical

Chemistry, 2006. 78(3): p. 646-654.

222. Nakamura, M., M. Shono, and K. Ishimura, Synthesis, Characterization, and Biological

Applications of Multifluorescent Silica Nanoparticles. Analytical Chemistry, 2007. 79(17):

p. 6507-6514.

223. Slowing, I.I., et al., Mesoporous silica nanoparticles as controlled release drug delivery

and gene transfection carriers. Advanced Drug Delivery Reviews, 2008. 60(11): p. 1278-

1288.

224. Chen, J.-F., et al., Preparation and characterization of porous hollow silica nanoparticles

for drug delivery application. Biomaterials, 2004. 25(4): p. 723-727.

225. Li, Z.Z., et al., Controlled release of avermectin from porous hollow silica nanoparticles:

Influence of shell thickness on loading efficiency, UV-shielding property and release.

Journal of Controlled Release, 2006. 111(1-2): p. 81-88.

226. Santra, S., Silica-based antibacterial and antifungal nanoformulation, Patent No.

WO/2010/068275, 2010.

References

215

227. Ebeling, W., Sorptive Dusts for Pest Control. Annual Review of Entomology, 1971. 16: p.

123-158.

228. Liang, R. and M.Z. Liu, Preparation of poly(acrylic acid-co-acrylamide)/kaolin and release

kinetics of urea from it. Journal of Applied Polymer Science, 2007. 106(5): p. 3007-3015.

229. Price, R. and A. Wagner, Method for treating agricultural crops using materials associated

with tubular carriers, Patent No. US 2008/0194406, 2008.

230. Pang, Y.-X. and X. Bao, Aluminium oxide nanoparticles prepared by water-in-oil

microemulsions. Journal Of Materials Chemistry, 2002. 12(12): p. 3699-3704.

231. Wu, J. and D. Xue, Progress of Science and Technology of ZnO as Advanced Material.

Science of Advanced Materials, 2011. 3(2): p. 127-149.

232. Mitchnick, M.A., D. Fairhurst, and S.R. Pinnell, Microfine zinc oxide (Z-Cote) as a

photostable UVA/UVB sunblock agent. Journal of the American Academy of

Dermatology, 1999. 40(1): p. 85-90.

233. Milani, N., et al., Dissolution Kinetics of Macronutrient Fertilizers Coated with

Manufactured Zinc Oxide Nanoparticles. Journal Of Agricultural And Food Chemistry,

2012. 60(16): p. 3991-3998.

234. Petersen, E.J., et al., Potential Release Pathways, Environmental Fate, And Ecological

Risks of Carbon Nanotubes. Environmental Science & Technology, 2011. 45(23): p.

9837-9856.

235. Pan, B. and B. Xing, Adsorption Mechanisms of Organic Chemicals on Carbon

Nanotubes. Environmental Science & Technology, 2008. 42(24): p. 9005-9013.

236. Biris, A. and M. Khodakovskaya, Method of using carbon nanotubes to affect seed

germination and plant growth, Patent No. WO/2011/059507, 2011.

237. Tripathi, S., S.K. Sonkar, and S. Sarkar, Growth stimulation of gram (Cicer arietinum)

plant by water soluble carbon nanotubes. Nanoscale, 2011. 3(3): p. 1176-1181.

238. Koelmans, A.A., et al., Black carbon: The reverse of its dark side. Chemosphere, 2006.

63(3): p. 365-377.

239. Baun, A., et al., Toxicity and bioaccumulation of xenobiotic organic compounds in the

presence of aqueous suspensions of aggregates of nano-C-60. Aquatic Toxicology, 2008.

86(3): p. 379-387.

240. Schwab, F., Ecotoxicology of Nanoparticles: Biota-Nanoparticle-Pollutant Interactions in

Aqueous Systems - Comparison of Black Carbon and Carbon Nanotubes, 2012, Phd

Thesis No. 20319, ETH Zurich.

241. Darlington, T.K., et al., Nanoparticle characteristics affecting environmental fate and

transport through soil. Environmental Toxicology and Chemistry, 2009. 28(6): p. 1191-

1199.

242. Fang, J., et al., Stability of titania nanoparticles in soil suspensions and transport in

saturated homogeneous soil columns. Environmental Pollution, 2009. 157(4): p. 1101-

1109.

Curriculum Vitae

216

243. Jaisi, D.P. and M. Elimelech, Single-walled carbon nanotubes exhibit limited transport in

soil columns. Environmental Science & Technology, 2009. 43: p. 9161-9166.

244. Wang, P., et al., Enhanced Environmental Mobility of Carbon Nanotubes in the Presence

of Humic Acid and Their Removal from Aqueous Solution. Small, 2008. 4(12): p. 2166-

2170.

245. Dubchak, S., et al., Influence of silver and titanium nanoparticles on arbuscular

mycorrhiza colonization and accumulation of radiocaesium in Helianthus annuus. Spanish

Journal of Agricultural Research, 2010. 8: p. S103-S108.

246. Ge, Y.G., J.P. Schimel, and P.A. Holden, Evidence for Negative Effects of TiO2 and ZnO

Nanoparticles on Soil Bacterial Communities. Environmental Science & Technology,

2011. 45(4): p. 1659-1664.

247. Yang, Y., et al., Relative Susceptibility and Transcriptional Response of Nitrogen Cycling

Bacteria to Quantum Dots. Environmental Science & Technology, 2012. 46(6): p. 3433-

3441.

248. Rico, C.M., et al., Interaction of Nanoparticles with Edible Plants and Their Possible

Implications in the Food Chain. Journal Of Agricultural And Food Chemistry, 2011. 59(8):

p. 3485-3498.

249. Kurepa, J., et al., Uptake and Distribution of Ultrasmall Anatase TiO2 Alizarin Red S

Nanoconjugates in Arabidopsis thaliana. Nano Letters, 2010. 10(7): p. 2296-2302.

250. Asli, S. and P.M. Neumann, Colloidal suspensions of clay or titanium dioxide

nanoparticles can inhibit leaf growth and transpiration via physical effects on root water

transport. Plant Cell And Environment, 2009. 32(5): p. 577-584.

251. Du, W.C., et al., TiO2 and ZnO nanoparticles negatively affect wheat growth and soil

enzyme activities in agricultural soil. Journal Of Environmental Monitoring, 2011. 13(4): p.

822-828.

252. Höck J., et al., Wegleitung zum Vorsorgeraster für Synthetische Nanomaterialien,

B.f.G.u.B.f. Umwelt, Editor 2011: Bern.

253. Nanosciences, nanotechnologies, materials & new production technologies (NMP). 2012

[cited 15.5.2012; Available from:

http://cordis.europa.eu/fp7/cooperation/nanotechnology_en.html.

254. European Parliament legislative resolution of 19 January 2012 on the Council position at

first reading with a view to the adoption of a regulation of the European Parliament and of

the Council concerning the making available on the market and use of biocidal products

(05032/2/2011 – C7-0251/2011 – 2009/0076(COD)), 2012, European Parliament.

255. Shrivastava, S. and D. Dash, Agrifood Nanotechnology: A tiny revolution in food and

agriculture. Journal Of Nano Research, 2009. 6: p. 1-14.

256. Gade, A., et al., Mycogenic metal nanoparticles: progress and applications.

Biotechnology Letters, 2010. 32(5): p. 593-600.

References

217

257. Magnuson, B.A., T.S. Jonaitis, and J.W. Card, A Brief Review of the Occurrence, Use,

and Safety of Food-Related Nanomaterials. Journal of Food Science, 2011. 76(6): p.

R126-R133.

258. Rai, M. and A. Ingle, Role of nanotechnology in agriculture with special reference to

management of insect pests. Applied Microbiology And Biotechnology, 2012. 94(2): p.

287-293.

259. Liu, F., et al., Porous hollow silica nanoparticles as controlled delivery system for water-

soluble pesticide. Materials Research Bulletin, 2006. 41(12): p. 2268-2275.

260. Liu, Y., Z. Tong, and R.K. Prud'homme, Stabilized polymeric nanoparticles for controlled

and efficient release of bifenthrin. Pest Management Science, 2008. 64(8): p. 808-812.

261. Boehm, A.L.L., R. Zerrouk, and H. Fessi, Poly epsilon-caprolactone nanoparticles

containing a poorly soluble pesticide: formulation and stability study. Journal Of

Microencapsulation, 2000. 17(2): p. 195-205.

262. Panichkin, L.A. and A.G. Raikova, Use of metal nano powder in presowing treatment of

seeds. Izvestiya Timiryazevskoi Sel'skokhozyaistvennoi Akademii, 2009(1): p. 59-65.

263. Song, S.L., et al., Stability of triazophos in self-nanoemulsifying pesticide delivery system.

Colloids And Surfaces A-Physicochemical And Engineering Aspects, 2009. 350(1-3): p.

57-62.

264. Kottegoda, N., et al., A green slow-release fertilizer composition based on urea-modified

hydroxyapatite nanoparticles encapsulated wood. Current Science, 2011. 101(1): p. 73-

78.

265. Gonzalez-Melendi, P., et al., Nanoparticles as smart treatment-delivery systems in plants:

Assessment of different techniques of microscopy for their visualization in plant tissues.

Annals of Botany, 2008. 101(1): p. 187-195.

266. Liu, Y., P. Laks, and P. Heiden, Controlled release of biocides in solid wood. III.

Preparation and characterization of surfactant-free nanoparticles. Journal Of Applied

Polymer Science, 2002. 86(3): p. 615-621.

267. Liu, X.-m., et al., Preparation and Testing of Cementing and Coating Nano-

Subnanocomposites of Slow/Controlled-Release Fertilizer. Agricultural Sciences in China,

2006. 5(9): p. 700-706.

268. Mondal, A., et al., Beneficial role of carbon nanotubes on mustard plant growth: an

agricultural prospect. Journal Of Nanoparticle Research, 2011. 13(10): p. 4519-4528.

269. bin Hussein, M.Z., et al., Controlled release of a plant growth regulator, alpha-

naphthaleneacetate from the lamella of Zn-Al-layered double hydroxide nanocomposite.

J. Controlled Release, 2002. 82(2-3): p. 417-427.

270. Spasova, M., et al., Electrospun biohybrid materials for plant biocontrol containing

chitosan and Trichoderma viride spores. Journal of Bioactive and Compatible Polymers,

2011. 26(1): p. 48-55.

271. Choi, K., S. Lee, and H. Choi, The liquid composition for promoting plant growth, which

includes nano-particle titanium dioxide, Patent No. WO/2003/059070, 2003.

Curriculum Vitae

218

272. Bignozzi, C., V. Dissette, and R. Della Valle, Products comprising an anti-microbial

composition based on titanium dioxide nanoparticles, Patent No. WO/2008/135093, 2008.

273. Young, D., et al., Fungicidal compositions including hydrazone derivatives and copper,

Patent No. WO/2010/083319, 2010.

274. Treseder, K.K. and M.D. Whiteside, Nanotechnological delivery of microbicides and other

substances, Patent No. WO/2011/031487, 2011.

275. Qiu, M. and Z. Li, Ultrafine powder or nano preparation of pyrethrins and preparation

process thereof, Patent No. CN1433697, 2003.

276. Dadajanov, T.K., et al., Biologically active nanochips for treating seeds of agricultural

plants e.g. rice seed comprises carrier; and biologically active component such as binder,

nanochip stabilizer, herbicide, plant-growth regulator, seed-germination stimulator, Patent

No. US2011000411-A1, 2011.

277. Maier, G. and G. Schneider, Verbesserung des Wachstums von Mikroorganismen, Patent

No. DE102009030121A1, 2010.

278. Park, H., Composition for controlling pathogenic microorganisms in plants, Patent No.

WO/2006/049379, 2006.

279. Shang, Q., H. Zheng, and T. Qi, Ivermectin water suspension nano capsule prepn and its

preparing method, Patent No. CN1491551, 2004.

280. Zheng, H. and Q. Shang, Water suspension acetamiprid nano capsule prepn and its

preparing method, Patent No. CN1491558, 2004.

281. Crooks, R., et al., Aqueous suspension of nanoparticles comprising an agrochemical

active ingredient, Patent No. WO/2002/082900, 2002.

282. Prasad, D.Y., Production of novel precision customized control release fertilizers, Patent

No. WO/2008/056234, 2008.

283. Greiner, A., et al., Invention concerning agricultural active substances, Patent No.

US8017061, 2011.

284. Koltzenburg, S., et al., Nanoparticulate active substance formulations, Patent No.

US7994227, 2011.

285. Greiner, A., et al., Device and use thereof for producing nanoscalar polymer fibres that

are used as supports for agricultural active ingredients, Patent No. WO/2005/115143,

2005.

286. Hayes, R.T., et al., Peham dendrimers for use in agriculture, Patent No.

WO/2011/053605, 2011.

287. Mahiran, B., et al., A herbicide formulation, Patent No. WO/2011/010910, 2011.

288. Wege, P.J., et al. A Microencapsulated Formulation of Lambda-Cyhalothrin. in 3rd

International Conference on Urban Pests. 1999.

289. Nowack, B., et al., Potential scenarios for nanomaterial release and subsequent alteration

in the environment. Environmental Toxicology and Chemistry, 2012. 31(1): p. 50-59.

290. Van Dijk, M.A., et al., Absorption and scattering microscopy of single metal nanoparticles.

Physical Chemistry Chemical Physics, 2006. 8(30): p. 3486-3495.

References

219

291. Mielke, R.E., et al., Differential growth of and nanoscale TiO2 accumulation in

Tetrahymena thermophila by direct feeding versus trophic transfer from Pseudomonas

aeruginosa. Applied and Environmental Microbiology, 2013. 79(18): p. 5616-5624.

292. Bondarenko, O., et al., Toxicity of Ag, CuO and ZnO nanoparticles to selected

environmentally relevant test organisms and mammalian cells in vitro: a critical review.

Archives of toxicology, 2013. 87(7): p. 1181-1200.

293. Juganson, K., et al., Extracellular conversion of silver ions into silver nanoparticles by

protozoan Tetrahymena thermophila. Environmental Science: Processes & Impacts,

2013. 15(1): p. 244-250.

294. Mortimer, M., K. Kasemets, and A. Kahru, Toxicity of ZnO and CuO nanoparticles to

ciliated protozoa Tetrahymena thermophila. Toxicology, 2010. 269(2-3): p. 182-9.

295. Petosa, A.R., et al., Aggregation and deposition of engineered nanomaterials in aquatic

environments: role of physicochemical interactions. Environ. Sci. Technol., 2010. 44(17):

p. 6532-49.

296. Jeong, S.W. and H. Kim, Filtration of fullerene and copper oxide nanoparticles using

surface-modified microfilters. Environ. Monit. Assess., 2014.

297. Morrow, J.B., R.D. Holbrook, and C.J. Zeissler, Quantum dot dissemination and behavior

in bacterial biofilms. Proceedings: NSTI-Nanotech 2006, 2006: p. 286-289.

298. Arregui, L., et al., Ciliate contributions to bioaggregation: laboratory assays with axenic

cultures of Tetrahymena thermophila. International Microbiology, 2007. 10(2): p. 91.

299. Hartmann, M., et al., Screening for and characterization of phospholipase A1

hypersecretory mutants of Tetrahymena thermophila. Applied Microbiology And

Biotechnology, 2000. 54(3): p. 390-396.

300. Madinger, C.L., et al., Constitutive secretion in Tetrahymena thermophila. Eukaryotic cell,

2010. 9(5): p. 674-681.

301. Leclerc, S. and K.J. Wilkinson, Bioaccumulation of Nanosilver by Chlamydomonas

reinhardtii- Nanoparticle or the Free Ion? Environmental Science & Technology, 2013.

48(1): p. 358-364.

302. Cytoviva - Hyperspectral imaging. 2012 [cited 05.06.2012; Available from:

http://www.cytoviva.com/.

303. Pokhrel, L.R., B. Dubey, and P.R. Scheuerman, Impacts of select organic ligands on the

colloidal stability, dissolution dynamics, and toxicity of silver nanoparticles. Environmental


304. Mortimer, M., A. Kahru, and V.I. Slaveykova, Uptake, localization and clearance of

quantum dots in ciliated protozoa Tetrahymena thermophila. Environmental Pollution,

2014. 190: p. 58-64.

305. Grabinski, C., J. Schlager, and S. Hussain, Hyperspectral Microscopy for

Characterization of Gold Nanoparticles in Biological Media and Cells for Toxicity

Assessment, in Nanomaterial Interfaces in Biology2013, Springer. p. 167-178.

Curriculum Vitae

220

306. Wen, Y., et al., Binding of cytoskeletal proteins with silver nanoparticles. RSC Advances,

2013. 3(44): p. 22002-22007.

307. Kadar, E., et al., Chemical interaction of atmospheric mineral dust-derived nanoparticles

with natural seawater—EPS and sunlight-mediated changes. Science Of The Total

Environment, 2014. 468: p. 265-271.

308. Zhang, S., et al., Aggregation, dissolution, and stability of quantum dots in marine

environments: importance of extracellular polymeric substances. Environmental Science

& Technology, 2012. 46(16): p. 8764-8772.

309. Shannahan, J.H., et al., Silver nanoparticle protein corona composition in cell culture

media. PloS one, 2013. 8(9): p. e74001.

310. Mortimer, M., et al., Exposure to CuO nanoparticles changes the fatty acid composition of

protozoa Tetrahymena thermophila. Environmental Science & Technology, 2011. 45(15):

p. 6617-6624.

311. von Moos, N. and V.I. Slaveykova, Oxidative stress induced by inorganic nanoparticles in

bacteria and aquatic microalgae-state of the art and knowledge gaps. Nanotoxicology,

2014. 8(6): p. 605-630.

312. Vetten, M., et al., Label-free in vitro toxicity and uptake assessment of citrate stabilised

gold nanoparticles in three cell lines. Particle and Fibre Toxicology, 2013. 10(1): p. 50.

313. Frankel, J., Cell biology of Tetrahymena thermophila. Methods Cell Biol, 1999. 62: p. 27-

125.

314. Ivask, A., et al., Toxicity Mechanisms in Escherichia coli Vary for Silver Nanoparticles and

Differ from Ionic Silver. Acs Nano, 2013. 8(1): p. 374-386.

315. Juganson, K., et al., Extracellular conversion of silver ions into silver nanoparticles by

protozoan Tetrahymena thermophila. Environ. Sci.: Processes Impacts, 2013. 15(1): p.

244-250.

316. Reidy, B., et al., Mechanisms of silver nanoparticle release, transformation and toxicity: A

critical review of current knowledge and recommendations for future studies and

applications. Materials, 2013. 6(6): p. 2295-2350.

317. von Moos, N. and V.I. Slaveykova, Oxidative stress induced by inorganic nanoparticles in

bacteria and aquatic microalgae--state of the art and knowledge gaps. Nanotoxicology,

2014. 8(6): p. 605-30.

318. Harrison, R.G. and G. Lunt, Membrane Components in Biological Membranes: Their

Structure and Function. 2nd ed1980, Glasgow, London: Blackie.

319. Yang, X.Y., et al., Mechanism of silver nanoparticle toxicity is dependent on dissolved

silver and surface coating in Caenorhabditis elegans. Environ. Sci. Technol., 2012. 46(2):

p. 1119-1127.

320. Ivask, A., et al., Mechanisms of toxic action of Ag, ZnO and CuO nanoparticles to

selected ecotoxicological test organisms and mammalian cells in vitro: A comparative

review. Nanotoxicology, Ahead of Print. doi:10.3109/17435390.2013.855831 .

References

221

321. Nanotechnology Consumer Products Inventory. Woodrow Wilson International Center for

Scholars, Nanotechnology Consumer Products Inventory,

http://www.nanotechproject.org/cpi/ 2014 [cited 28.1.2014; Available from:

http://www.nanotechproject.org/inventories/consumer/search/.

322. Gottschalk, F., T. Sun, and B. Nowack, Environmental concentrations of engineered

nanomaterials: Review of modeling and analytical studies. Environmental Pollution, 2013.

181: p. 287-300.

323. Khodakovskaya, M.V., et al., Carbon Nanotubes as Plant Growth Regulators: Effects on

Tomato Growth, Reproductive System, and Soil Microbial Community. Small, 2013. 9(1):

p. 115-123.

324. Gogos, A., K. Knauer, and T.D. Bucheli, Nanomaterials in Plant Protection and

Fertilization: Current State, Foreseen Applications, and Research Priorities. Journal Of

Agricultural And Food Chemistry, 2012. 60(39): p. 9781-9792.

325. Tong, Z., et al., Response of Soil Microorganisms to As-Produced and Functionalized

Single-Wall Carbon Nanotubes (SWNTs). Environmental Science & Technology, 2012.

46(24): p. 13471-13479.

326. Canas, J.E., et al., Effects of functionalized and nonfunctionalized single-walled carbon

nanotubes on root elongation of select crop species. Environmental Toxicology and

Chemistry, 2008. 27(9): p. 1922-1931.

327. Colbeck, I., B. Atkinson, and Y. Johar, The morphology and optical properties of soot

produced by different fuels. Journal of Aerosol Science, 1997. 28(5): p. 715-723.

328. Chen, B. and J.P. Selegue, Separation and Characterization of Single-Walled and

Multiwalled Carbon Nanotubes by Using Flow Field-Flow Fractionation. Analytical

Chemistry, 2002. 74(18): p. 4774-4780.

329. Tagmatarchis, N., et al., Separation and purification of functionalised water-soluble multi-

walled carbon nanotubes by flow field-flow fractionation. Carbon, 2005. 43(9): p. 1984-

1989.

330. Chun, J., et al., Size Separation of Single-Wall Carbon Nanotubes by Flow-Field Flow

Fractionation. Analytical Chemistry, 2008. 80(7): p. 2514-2523.

331. Gigault, J., et al., Accurate determination of the length of carbon nanotubes using multi-

angle light scattering. Microchimica Acta, 2011. 175(3-4): p. 265-271.

332. Gigault, J., B. Grassl, and G. Lespes, Multi-wall carbon nanotube aqueous dispersion

monitoring by using A4F-UV-MALS. Analytical And Bioanalytical Chemistry, 2011.

401(10): p. 3345-3353.

333. Martin, M. and P.S. Williams, Theoretical basis of field-flow fractionation, in Theoretical

Advancement in Chromatography and Related Separation Techniques1992, Springer. p.

513-580.

334. Kammer, F.v.d., Characterization of Environmental Colloids applying Field-Flow

Fractionation - Multi Detection Analysis with Emphasis on Light Scattering Techniques,

2004, PhD Thesis, TU Hamburg-Harburg.

Curriculum Vitae

222

335. Schwab, F., et al., Are carbon nanotube effects on green algae caused by shading and

agglomeration? Environmental Science & Technology, 2011. 45(14): p. 6136-6144.

336. Schwab, F., et al., Sorption kinetics and equilibrium of the herbicide diuron to carbon

nanotubes or soot in absence and presence of algae. Environmental Pollution, 2014. 192:

p. 147-153.

337. Hammes, K., et al., Comparison of quantification methods to measure fire-derived

(black/elemental) carbon in soils and sediments using reference materials from soil,

water, sediment and the atmosphere. Global Biogeochemical Cycles, 2007. 21(3): p.

GB3016.

338. Arganda-Carreras, I., et al., 3D reconstruction of histological sections: Application to

mammary gland tissue. Microscopy research and technique, 2010. 73(11): p. 1019-1029.

339. Doube, M., et al., BoneJ: Free and extensible bone image analysis in ImageJ. Bone,

2010. 47(6): p. 1076-1079.

340. Polder, G., H.L.E. Hovens, and A.J. Zweers, Measuring shoot length of submerged

aquatic plants using graph analysis. Proceedings of the ImageJ User and Developer

Conference, 2010. 27-29 p. 172-177.

341. Kaegi, R., Chemical and morphological analysis of airborne particles at a tunnel

construction site. Journal of Aerosol Science, 2004. 35(5): p. 621-632.

342. Keith, L.H., et al., Principles of environmental analysis. Analytical Chemistry, 1983.

55(14): p. 2210-2218.

343. Plata, D.L., P.M. Gschwend, and C.M. Reddy, Industrially synthesized single-walled

carbon nanotubes: Compositional data for users, environmental risk assessments, and

source apportionment. Nanotechnology, 2008. 19(18): p. 185706.

344. Miyata, Y., K. Mizuno, and H. Kataura, Purity and defect characterization of single-wall

carbon nanotubes using Raman spectroscopy. Journal of Nanomaterials, 2011. 2011: p.

1-7.

345. Phelan Jr, F.R. and B.J. Bauer, Comparison of steric effects in the modeling of spheres

and rodlike particles in field-flow fractionation. Chemical Engineering Science, 2009.

64(8): p. 1747-1758.

346. Matsuoka, K. and Y. Moroi, Micelle formation of sodium deoxycholate and sodium

ursodeoxycholate (Part 1). Biochimica et Biophysica Acta, 2002. 1580: p. 189-199.

347. Arnold, M.S., et al., Sorting carbon nanotubes by electronic structure using density

differentiation. Nature Nanotechnology, 2006. 1(1): p. 60-65.

348. Hassellöv, M., F.v. der Kammer, and R. Beckett, Characterisation of Aquatic Colloids and

Macromolecules by Field-Flow Fractionation, in Environmental Colloids and

Particles2007, John Wiley & Sons, Ltd. p. 223-276.

349. Musumeci, A.W., et al., Thermal decomposition and electron microscopy studies of

single-walled carbon nanotubes. Journal of Thermal Analysis and Calorimetry, 2007.

88(3): p. 885-891.

References

223

350. Reed, R.B., et al., Detection of single walled carbon nanotubes by monitoring embedded

metals. Environmental Science: Processes & Impacts, 2013. 15(1): p. 204-213.

351. Kuo, D.T.F., J.B. Vander Sande, and P.M. Gschwend, Characterization of black carbon in

geosorbents at the nanometer scale by STEM-EDX elemental mapping. Organic

Geochemistry, 2013. 56(0): p. 81-93.

352. BUWAL, Nationales Bodenbeobachtungsnetz: Messresultate 1985-1991, (Schriftenreihe

Umwelt Nr. 200).

353. Wang, S., et al., Statistical characterization of single-wall carbon nanotube length

distribution. Nanotechnology, 2006. 17(3): p. 634.

354. Krause, B., R. Boldt, and P. Pötschke, A method for determination of length distributions

of multiwalled carbon nanotubes before and after melt processing. Carbon, 2011. 49(4):

p. 1243-1247.

355. Laurent, C., E. Flahaut, and A. Peigney, The weight and density of carbon nanotubes

versus the number of walls and diameter. Carbon, 2010. 48(10): p. 2994-2996.

356. Soong, C., P. Woo, and D. Hoyle, Contamination cleaning of TEM/SEM samples with the

ZONE cleaner. Microscopy Today, 2012. 20(06): p. 44-48.

357. Kah, M., et al., Nanopesticides: State of Knowledge, Environmental Fate, and Exposure

Modeling. Critical Reviews In Environmental Science And Technology, 2012. 43(16): p.

1823-1867.

358. Chen, G., X. Liu, and C. Su, Transport and retention of TiO2 rutile nanoparticles in

saturated porous media under low-ionic-strength conditions: Measurements and

mechanisms. Langmuir, 2011. 27(9): p. 5393-5402.

359. Solovitch, N., et al., Concurrent aggregation and deposition of TiO2 nanoparticles in a

sandy porous media. Environmental Science & Technology, 2010. 44(13): p. 4897-4902.

360. Servin, A.D., et al., Synchrotron micro-XRF and micro-XANES confirmation of the uptake

and translocation of TiO2 nanoparticles in cucumber (Cucumis sativus) plants.


361. Larue, C., et al., Comparative Uptake and Impact of TiO2 Nanoparticles in Wheat and

Rapeseed. Journal of Toxicology and Environmental Health, Part A, 2012. 75(13-15): p.

722-734.

362. Liu, Q.L., et al., Carbon Nanotubes as Molecular Transporters for Walled Plant Cells.

Nano Letters, 2009. 9(3): p. 1007-1010.

363. Kasel, D., et al., Limited transport of functionalized multi-walled carbon nanotubes in two

natural soils. Environmental Pollution, 2013. 180(0): p. 152-158.

364. Tian, Y., et al., Deposition and transport of functionalized carbon nanotubes in water-

saturated sand columns. Journal of Hazardous Materials, 2012. 213–214(0): p. 265-272.

365. Moll, J., et al., Titanium dioxide nanoparticles and carbon nanotubes in a soil mesocosm:

effects on red clover and its symbiotic microorganisms in a soil system. In preparation for

submission to Environmental Pollution, 2015.

Curriculum Vitae

224

366. Moll, J., et al., Titanium dioxide nanoparticles in mesocosms: effects on wheat and soil

microbes. In preparation for submission, 2015.

367. Gogos, A., et al., Capabilities of asymmetric flow field-flow fractionation coupled to multi-

angle light scattering to detect carbon nanotubes in soot and soil. Environmental Science:

Nano, 2014. 1(6): p. 584-594.

368. Dunphy Guzman, K.A., M.P. Finnegan, and J.F. Banfield, Influence of surface potential

on aggregation and transport of titania nanoparticles. Environmental Science &

Technology, 2006. 40(24): p. 7688-7693.

369. Dumon, J.C. and W.H.O. Ernst, Titanium in Plants. Journal of Plant Physiology, 1988.

133(2): p. 203-209.

370. Larue, C., et al., Accumulation, translocation and impact of TiO2 nanoparticles in wheat

(Triticum aestivum spp.): Influence of diameter and crystal phase. Science Of The Total

Environment, 2012. 431(0): p. 197-208.

371. Gierlinger, N. and M. Schwanninger, The potential of Raman microscopy and Raman

imaging in plant research. Spectroscopy: An International Journal, 2007. 21(2): p. 69-89.

372. de Neergaard, A., et al., Decomposition of white clover (Trifolium repens) and ryegrass

(Lolium perenne) components: C and N dynamics simulated with the DAISY soil organic

matter submodel. European Journal of Agronomy, 2002. 16(1): p. 43-55.

373. Serag, M.F., et al., Trafficking and Subcellular Localization of Multiwalled Carbon

Nanotubes in Plant Cells. Acs Nano, 2011. 5(1): p. 493-499.

374. Lucivero, F., T. Swierstra, and M. Boenink, Assessing expectations: towards a toolbox for

an ethics of emerging technologies. NanoEthics, 2011. 5(2): p. 129-141.

375. Kah, M., et al., Nanopesticides: State of Knowledge, Environmental Fate, and Exposure

Modeling. Critical Reviews In Environmental Science And Technology, 2013. 43(16): p.

1823-1867.

376. Feng, B.H. and Z.Y. Zhang, Carboxymethy chitosan grafted ricinoleic acid group for

nanopesticide carriers. Advanced Materials Research, 2011. 236: p. 1783-1788.

377. Cota‐Arriola, O., et al., Controlled release matrices and micro/nanoparticles of chitosan

with antimicrobial potential: development of new strategies for microbial control in

agriculture. Journal of the Science of Food and Agriculture, 2013. 93(7): p. 1525-1536.

378. Zhang, J., et al., Construction of novel amphiphilic chitosan copolymer nanoparticles for

chlorpyrifos delivery. Journal of Polymer Research, 2013. 20(3): p. 1-11.

379. Sharp, R.G., A review of the applications of chitin and its derivatives in agriculture to

modify plant-microbial interactions and improve crop yields. Agronomy, 2013. 3(4): p.

757-793.

380. De, A., et al., Targeted Delivery of Pesticides Using Biodegradable Polymeric

Nanoparticles2014: Springer.

381. Nadiminti, P.P., et al., Nanostructured liquid crystalline particles as an alternative delivery

vehicle for plant agrochemicals. ACS applied materials & interfaces, 2013. 5(5): p. 1818-

1826.

References

225

382. Lim, C.J., et al., Physicochemical characterization and formation of glyphosate-laden

nano-emulsion for herbicide formulation. Industrial Crops and Products, 2012. 36(1): p.

607-613.

383. Chaw Jiang, L., et al., Green nano-emulsion intervention for water-soluble glyphosate

isopropylamine (IPA) formulations in controlling Eleusine indica (E. indica). Pesticide

Biochemistry and Physiology, 2012. 102(1): p. 19-29.

384. Cai, D., et al., Controlling Pesticide Loss by Natural Porous Micro/Nano Composites:

Straw Ash-Based Biochar and Biosilica. ACS applied materials & interfaces, 2013. 5(18):

p. 9212-9216.

385. Hu, M., et al., Dark-field microscopy studies of single metal nanoparticles: understanding

the factors that influence the linewidth of the localized surface plasmon resonance.

Journal Of Materials Chemistry, 2008. 18(17): p. 1949-1960.

386. Swiss Federal Office for Agriculture (FOAG). [cited 12.01.2014; Available from:

http://www.blw.admin.ch/themen/00011/00075/00222/index.html?lang=de.

387. Éléments issus des déclarations des substances à l’état nanoparticulaire – exercice

2014. 2014 [cited 12.01.2015; Available from: http://www.developpement-

durable.gouv.fr/IMG/pdf/rapport-nano-2014.pdf.

388. Baun, A., et al., Toxicity and bioaccumulation of xenobiotic organic compounds in the

presence of aqueous suspensions of aggregates of nano-C60. Aquatic Toxicology, 2008.

86(3): p. 379-387.

389. http://www.hausserscientific.com/products/petroff_hausser_counter.html. [cited

13.01.2015].

390. Mudalige, T.K., et al., Simple Functionalization Strategies for Enhancing Nanoparticle

Separation and Recovery with Asymmetric Flow Field Flow Fractionation. Analytical

Chemistry, 2015.

391. Sadezky, A., et al., Raman microspectroscopy of soot and related carbonaceous

materials: spectral analysis and structural information. Carbon, 2005. 43(8): p. 1731-

1742.

392. Beech, J.P., et al., Sorting cells by size, shape and deformability. Lab on a Chip, 2012.

12(6): p. 1048-1051.

393. Mao, H., et al., Microfluidic Surface‐Enhanced Raman Scattering Sensors Based on

Nanopillar Forests Realized by an Oxygen‐Plasma‐Stripping‐of‐Photoresist Technique.

Small, 2014. 10(1): p. 127-134.

394. Dolcater, D., J. Syers, and M. Jackson, Titanium as free oxide and substituted forms in

kaolinites and other soil minerals. Clays Clay Miner, 1970. 18: p. 71-79.

395. Bickley, R.I., et al., A structural investigation of titanium dioxide photocatalysts. Journal of

Solid State Chemistry, 1991. 92(1): p. 178-190.

Curriculum Vitae

226

396. Brydson, R., et al., Electron energy loss and X-ray absorption spectroscopy of rutile and

anatase: a test of structural sensitivity. Journal of Physics: Condensed Matter, 1989. 1(4):

p. 797.

397. Sartori, A., et al., Correlative microscopy: bridging the gap between fluorescence light

microscopy and cryo-electron tomography. Journal of structural biology, 2007. 160(2): p.

135-145.

398. Jiruse, J., et al., FIB-SEM Instrument with Integrated Raman Spectroscopy for Correlative

Microscopy. Microscopy and Microanalysis, 2014. 20(S3): p. 990-991.

diss deposit final ohnecv - research collection47855/et… · diss. eth no. 22589 engineered...

Documents